# BulkTranscripts

> BulkTranscripts downloads every available transcript from a YouTube channel, playlist, or video in one job and
> exports them as TXT, Markdown, SRT, VTT, CSV, JSON, or a single combined AI-ready file built for
> ChatGPT, Claude, and NotebookLM. It also exposes the same engine as a REST API, an MCP server,
> and an installable agent skill, so AI agents can fetch transcripts, search YouTube, list
> channels/playlists, and track new uploads directly.

Bulk extraction is the point: most tools handle one video at a time, which does not help when you
want a whole archive. The web app needs no account for small runs: anonymous visitors get a free allowance; runs over 30 videos ask for a Google sign-in. The
REST API needs an API key, which is free: sign in with Google at /app, open the MCP & API tab, and
create one under API keys (starts with bt_ak_, 30 free credits, no card). Send it as
Authorization: Bearer on every /api/v1/* request; a request with no key returns 401
missing_api_key. The MCP connector takes one OAuth sign-in (Google) instead, and an API key also
works on /mcp for clients that cannot run OAuth. Credits follow the account, so a one-time credit
pack tops up the same balance and the key never changes. A license key from checkout is also an
API key.

## Core
- [App](https://bulktranscripts.co/app): paste a URL and extract
- [API & MCP docs](https://bulktranscripts.co/docs): REST API, MCP server, agent skill
- [Workflows](https://bulktranscripts.co/workflows): copy-paste recipes for study notes, channel monitoring, RAG ingestion, show notes
- [OpenAPI spec](https://bulktranscripts.co/openapi.json): machine-readable API description
- [Guides](https://bulktranscripts.co/how-to): per-channel and per-format guides

## For AI agents (recommended)
- [MCP server](https://bulktranscripts.co/mcp): streamable HTTP, OAuth sign-in (Google). Tools: get_transcript, get_transcripts (batch of 20), search_youtube, search_channel, get_channel_videos, get_playlist_videos, get_latest_videos (free), start_bulk_extract (whole channel or playlist, up to 1,000 videos), get_bulk_status (free), list_bulk_results (free). Connect: claude mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp
- [MCP server guide](https://bulktranscripts.co/youtube-mcp-server): what each tool returns and costs
- [Agent skill](https://bulktranscripts.co/skill.md): save as SKILL.md; teaches the whole REST API
- [Transcript endpoint](https://bulktranscripts.co/api/v1/transcript?video=<url-or-id>): transcript JSON (1 credit on first library addition; repeat reads are free)
- [Search YouTube](https://bulktranscripts.co/api/v1/search?q=<query>): 1 credit
- [Search inside a channel](https://bulktranscripts.co/api/v1/channel/search?channel=@handle&q=<query>): 1 credit
- [List channel videos](https://bulktranscripts.co/api/v1/channel/videos?channel=@handle): 1 credit
- [Latest uploads](https://bulktranscripts.co/api/v1/channel/latest?channel=@handle): newest videos, FREE; poll it to detect new uploads
- [List playlist videos](https://bulktranscripts.co/api/v1/playlist/videos?playlist=<id>): 1 credit
- [Start a bulk job](https://bulktranscripts.co/api/v1/bulk): POST a channel or playlist URL; up to 1,000 videos, listing free, 1 credit per new transcript
- [Bulk job status](https://bulktranscripts.co/api/v1/bulk/<run_id>): free; poll at the pace it asks for
- [Bulk job results](https://bulktranscripts.co/api/v1/bulk/<run_id>/results): free; pages through each video's outcome
- [Account balance](https://bulktranscripts.co/api/v1/account): free
- [Full text of the key pages](https://bulktranscripts.co/llms-full.txt): docs, MCP, API and integration pages concatenated

## Tools and landing pages
- [YouTube to transcript](https://bulktranscripts.co/youtube-to-transcript): Turn a YouTube video into a clean, readable transcript with optional timestamps. Read it on the page, copy it, or download it free. No sign-up required.
- [YouTube channel transcript downloader](https://bulktranscripts.co/youtube-channel-transcript-downloader): Download the transcript of every video on a YouTube channel in one job. Export TXT, Markdown, SRT, VTT, CSV, JSON, or one combined AI-ready file. Start with 30 free transcripts, no credit card.
- [YouTube playlist transcript downloader](https://bulktranscripts.co/youtube-playlist-transcript-downloader): Download transcripts for every video in a YouTube playlist at once. Export to TXT, Markdown, SRT, VTT, CSV, JSON, or one combined AI-ready file. Start with 30 free transcripts, no credit card.
- [YouTube transcript downloader](https://bulktranscripts.co/youtube-transcript-downloader): Download YouTube transcripts from one video, a playlist or a whole channel. Export TXT, Markdown, SRT, VTT, CSV or JSON. First 30 free, no card needed.
- [Download all transcripts from a YouTube channel](https://bulktranscripts.co/download-all-youtube-channel-transcripts): Download every transcript from a YouTube channel in one job: up to 1,000 videos in about 7 minutes, as a ZIP or one combined file. 30 transcripts free.
- [Get transcripts for many YouTube links at once](https://bulktranscripts.co/bulk-youtube-links-to-transcripts): Paste a list of YouTube video links, one per line, and get every transcript in one job. Up to 1,000 links, duplicates removed, one ZIP or one combined file. 30 free transcripts.
- [YouTube transcripts for NotebookLM](https://bulktranscripts.co/youtube-transcripts-for-notebooklm): NotebookLM adds one YouTube video at a time. Export an entire channel or playlist as a single source file you upload once, formatted with the metadata NotebookLM needs to cite.
- [YouTube channel to text](https://bulktranscripts.co/youtube-channel-to-text): Convert an entire YouTube channel to clean, readable text. Paragraphs instead of caption fragments, exported as TXT, Markdown, or one combined document.
- [YouTube subtitle downloader](https://bulktranscripts.co/youtube-subtitle-downloader): Download YouTube subtitles as SRT, VTT, TXT, Markdown, CSV or JSON. Works on one video or a whole channel. Start with 30 free transcripts; no extension.
- [Bulk YouTube subtitle downloader](https://bulktranscripts.co/bulk-youtube-subtitle-downloader): Download subtitles from every video on a YouTube channel or playlist in one job. Get a ZIP of SRT, VTT or text files. Up to 1,000 videos, free to start, with 30 free transcript extractions and no credit card.
- [YouTube to JSON](https://bulktranscripts.co/youtube-to-json): Export any YouTube video, playlist or channel to JSON: id, title, channel, URL, duration, date and the full timed transcript. No code, no API key. 30 free.
- [YouTube transcript with timestamps](https://bulktranscripts.co/youtube-transcript-with-timestamps): Download a YouTube transcript with timestamps intact. Choose SRT, VTT, CSV or JSON to keep cue timings, or strip them for clean reading text. Start with 30 free transcripts, no credit card.
- [YouTube transcript API](https://bulktranscripts.co/youtube-transcript-api): One HTTP call returns timestamped JSON for any video, channel or playlist. We run the residential proxies, so requests don't get blocked. Free key, 30 credits.
- [YouTube MCP server for transcripts, search and channels](https://bulktranscripts.co/youtube-mcp-server): A hosted YouTube MCP server: transcripts, search, channels and playlists for Claude, ChatGPT, Cursor and VS Code. Nothing to install, 30 free credits.
- [Give your AI agent access to YouTube transcripts](https://bulktranscripts.co/youtube-transcript-agent-skill): A SKILL.md on GitHub that lets Claude Code, Codex, Cursor, or OpenClaw fetch YouTube transcripts, search channels, and list playlists. Free key with 30 credits.
- [YouTube transcripts for AI agents](https://bulktranscripts.co/youtube-api-for-ai-agents): Feed YouTube into your AI agents: a transcript API with a shared cache, an MCP server, and an installable agent skill. Built for RAG pipelines, summarizers, and autonomous research agents.
- [How to download or copy a YouTube transcript](https://bulktranscripts.co/how-to-download-youtube-transcript): Two ways to get a YouTube transcript: YouTube's built-in Show transcript panel for copying on desktop, and a free tool for downloading a TXT, SRT or Markdown file on any device, including phones where the panel is missing.

## Integrations
- [Claude Code](https://bulktranscripts.co/integrations/claude-code-youtube): Connect Claude Code to YouTube in one command: transcripts, channel search, playlists, and upload tracking via the BulkTranscripts MCP server or agent skill. One sign-in, 30 free credits.
- [Codex](https://bulktranscripts.co/integrations/codex-youtube): Give OpenAI Codex CLI YouTube access: add the BulkTranscripts MCP server to config.toml or install the agent skill. Transcripts, channel search, playlists. One sign-in or license key.
- [Cursor](https://bulktranscripts.co/integrations/cursor-youtube): Add the BulkTranscripts YouTube MCP server to Cursor with one click or one JSON block: transcripts, channel research, and upload tracking in Cursor's agent. One sign-in, 30 free credits.
- [Grok Bot](https://bulktranscripts.co/integrations/grok-bot-youtube-transcripts): Paste a YouTube link into Grok and it answers from the title, not what was said. Here's how to make it read the real video: one link or a whole channel.
- [ChatGPT](https://bulktranscripts.co/integrations/chatgpt-youtube): ChatGPT can't watch YouTube, but it can read it. Add one connector and it pulls the transcript of any video, playlist or channel. Google sign-in, 30 free.
- [VS Code](https://bulktranscripts.co/integrations/vscode-youtube): Add the BulkTranscripts YouTube MCP server to VS Code's agent mode via mcp.json: transcripts, channel search, and playlists for Copilot. One JSON block plus a sign-in.
- [n8n](https://bulktranscripts.co/integrations/n8n-youtube-transcripts): Build n8n workflows on YouTube transcripts: the BulkTranscripts node (verified by n8n), exact HTTP Request node settings, a free new-upload trigger pattern, and the MCP Client option for AI Agent nodes.
- [Make](https://bulktranscripts.co/integrations/make-youtube-transcripts): Pull YouTube transcripts into Make scenarios with the HTTP module: exact settings, a free daily upload-check pattern, and routing transcripts to Sheets, Docs, or AI modules.
- [Gemini CLI](https://bulktranscripts.co/integrations/gemini-cli-youtube): Give Gemini CLI YouTube access in one command: install the BulkTranscripts extension or add the MCP server. Transcripts, channel search and playlists, with a Google sign-in and 30 free credits.
- [Claude Cowork](https://bulktranscripts.co/integrations/claude-cowork-youtube): Add YouTube transcripts to Claude Cowork as a custom connector: paste one URL, sign in with Google, and Cowork can read videos, channels and playlists. 30 free credits.
- [Perplexity Computer](https://bulktranscripts.co/integrations/perplexity-computer-youtube): Connect Perplexity Computer to YouTube transcripts with a custom remote connector: one URL, Google sign-in, 30 free credits. Read videos, channels and playlists inside Computer tasks.
- [Manus](https://bulktranscripts.co/integrations/manus-youtube): Add YouTube transcripts to Manus as a custom MCP connector: transcripts, channel research and playlists for Manus tasks. Google sign-in or an API key, 30 free credits.
- [OpenClaw](https://bulktranscripts.co/integrations/openclaw-youtube): Give OpenClaw YouTube access: add the BulkTranscripts MCP server with OAuth or install the skill from ClawHub. Transcripts, channel search and playlists, 30 free credits.
- [Hermes Agent](https://bulktranscripts.co/integrations/hermes-agent-youtube): Connect Hermes Agent by Nous Research to YouTube transcripts with one command: the BulkTranscripts MCP server with OAuth or an API key. Videos, channels and playlists, 30 free credits.
- [Kimi Code](https://bulktranscripts.co/integrations/kimi-code-youtube): Give Kimi Code, Moonshot AI's terminal agent, YouTube access: one mcp.json entry for the BulkTranscripts MCP server, a Google sign-in, 30 free credits.
- [NanoClaw](https://bulktranscripts.co/integrations/nanoclaw-youtube): Connect NanoClaw to YouTube transcripts: add the BulkTranscripts MCP server to a group with ncl and keep the API key in the vault. Videos, channels, playlists.
- [Paperclip](https://bulktranscripts.co/integrations/paperclip-youtube): Give your Paperclip agents YouTube transcripts: add the BulkTranscripts skill from the Skills page, or connect the MCP server in the runtime Paperclip launches.

## Workflows
- [YouTube playlist → study notes](https://bulktranscripts.co/workflows/playlist-study-notes): Convert a course or lecture playlist into per-video study notes with an AI assistant: transcripts in playlist order, summaries, key concepts, and definitions ready for Notion or Markdown.
- [Competitor channels → weekly brief](https://bulktranscripts.co/workflows/competitor-channel-brief): Track what competitor channels publish and get a briefing on what matters: free new-upload checks, transcripts only for relevant videos, quotes with timestamps.
- [Playlist → RAG corpus](https://bulktranscripts.co/workflows/playlist-to-rag-corpus): Build a retrieval corpus from a YouTube playlist: timestamped transcript segments as natural chunk boundaries, metadata for citations, one API loop to ingest everything.
- [YouTube video → blog post](https://bulktranscripts.co/workflows/video-to-blog-post): Draft a blog post from a YouTube video using its real transcript: accurate quotes, the speaker's actual structure, and none of the hallucinated paraphrasing you get from summarizing thumbnails.
- [New uploads → summarized alerts](https://bulktranscripts.co/workflows/new-upload-alerts): Get notified with a summary when channels you follow publish: a free RSS-backed check, transcript fetch only for new videos, and a digest to Slack, email, or Notion.
- [Channel archive → cited quotes](https://bulktranscripts.co/workflows/research-quotes-timestamps): Search inside a creator's archive for what they said about a topic, and cite it properly: exact quotes, video links, and second-accurate timestamps.
- [Episode → show notes & chapters](https://bulktranscripts.co/workflows/podcast-show-notes): Turn a podcast episode on YouTube into show notes: chapter markers from real timestamps, guest quotes, links mentioned, and a description that matches what was actually said.
- [Whole channel → knowledge base](https://bulktranscripts.co/workflows/channel-knowledge-base): Turn a channel's entire archive (tutorials, talks, internal trainings) into a searchable text knowledge base your team and your AI tools can actually query.
- [One video → a week of posts](https://bulktranscripts.co/workflows/video-to-social-posts): Repurpose one YouTube video into an X thread and a LinkedIn post using the actual transcript: the speaker's real lines, not a vague summary of the thumbnail.
- [Lectures → flashcards & quizzes](https://bulktranscripts.co/workflows/lecture-flashcards-quiz): Generate spaced-repetition flashcards and self-test quizzes from YouTube lectures: transcript-grounded Q&A pairs, Anki-ready output, timestamps back to the explanation.

## Comparisons (pricing snapshots dated on each page)
- [BulkTranscripts vs TranscriptAPI](https://bulktranscripts.co/compare/bulktranscripts-vs-transcriptapi): TranscriptAPI is a developer API billed monthly. BulkTranscripts adds a bulk web app, one-time credits that never expire, and the same REST, MCP and agent skill surface. Pricing, limits and fit, side by side.
- [BulkTranscripts vs youtube-transcript.io](https://bulktranscripts.co/compare/bulktranscripts-vs-youtube-transcript-io): youtube-transcript.io sells monthly subscriptions with a bulk feature marked experimental. BulkTranscripts is built around bulk: whole channels, one-time credits that never expire, and an MCP server for agents. Compared honestly.
- [BulkTranscripts vs Supadata](https://bulktranscripts.co/compare/bulktranscripts-vs-supadata): Supadata is a multi-platform scraping API with monthly credits. BulkTranscripts is YouTube-first with a bulk web app, one-time credits and a hosted MCP server. Pricing, credit rules and fit, side by side.
- [Tactiq YouTube transcript generator vs BulkTranscripts](https://bulktranscripts.co/compare/bulktranscripts-vs-tactiq): Tactiq's free YouTube transcript generator handles one video at a time and downloads TXT. BulkTranscripts does whole channels and playlists, six export formats, and an API. What each is good for, with dated facts from Tactiq's site.
- [NoteGPT YouTube transcript generator vs BulkTranscripts](https://bulktranscripts.co/compare/bulktranscripts-vs-notegpt): NoteGPT bundles YouTube transcripts with AI summaries, notes and a monthly subscription from $9. BulkTranscripts sells one-time credits, exports six formats and exposes an API and MCP server. Limits, prices and fit, checked on notegpt.io.
- [Best YouTube MCP servers in 2026](https://bulktranscripts.co/best-youtube-mcp-servers): Nine YouTube MCP servers for Claude, ChatGPT, Cursor and VS Code, compared: hosted vs self-hosted, transcripts only vs search and channel listings, and what each costs to run. GitHub stars as of 2026-09-08.
- [Best YouTube transcript tools in 2026](https://bulktranscripts.co/best-youtube-transcript-tools): Ten ways to get YouTube transcripts in 2026, compared on bulk channel support, API and MCP access, formats and real prices: BulkTranscripts, TranscriptAPI, Supadata, youtube-transcript.io, NoteGPT, Tactiq, Apify, yt-dlp, youtube-transcript-api and DownSub. Prices checked 2026-10-03.

## Guides
- [Huberman Lab transcripts](https://bulktranscripts.co/how-to/huberman-lab-transcripts): download every available transcript from Huberman Lab in bulk
- [Lex Fridman Podcast transcripts](https://bulktranscripts.co/how-to/lex-fridman-podcast-transcripts): download every available transcript from Lex Fridman Podcast in bulk
- [The Joe Rogan Experience transcripts](https://bulktranscripts.co/how-to/joe-rogan-experience-transcripts): download every available transcript from The Joe Rogan Experience in bulk
- [the All-In Podcast transcripts](https://bulktranscripts.co/how-to/all-in-podcast-transcripts): download every available transcript from the All-In Podcast in bulk
- [MKBHD transcripts](https://bulktranscripts.co/how-to/mkbhd-video-transcripts): download every available transcript from MKBHD in bulk
- [TED transcripts](https://bulktranscripts.co/how-to/ted-talk-transcripts): download every available transcript from TED in bulk
- [Y Combinator transcripts](https://bulktranscripts.co/how-to/y-combinator-transcripts): download every available transcript from Y Combinator in bulk
- [a university lecture playlist transcripts](https://bulktranscripts.co/how-to/lecture-playlist-transcripts): download every available transcript from a university lecture playlist in bulk
- [a conference playlist transcripts](https://bulktranscripts.co/how-to/conference-talk-transcripts): download every available transcript from a conference playlist in bulk
- [20VC transcripts](https://bulktranscripts.co/how-to/20vc-podcast-transcripts): download every available transcript from 20VC in bulk
- [the Shawn Ryan Show transcripts](https://bulktranscripts.co/how-to/shawn-ryan-show-transcripts): download every available transcript from the Shawn Ryan Show in bulk
- [Club Shay Shay transcripts](https://bulktranscripts.co/how-to/club-shay-shay-transcripts): download every available transcript from Club Shay Shay in bulk
- [Jocko Podcast transcripts](https://bulktranscripts.co/how-to/jocko-podcast-transcripts): download every available transcript from Jocko Podcast in bulk
- [Call Her Daddy transcripts](https://bulktranscripts.co/how-to/call-her-daddy-transcripts): download every available transcript from Call Her Daddy in bulk
- [Earn Your Leisure transcripts](https://bulktranscripts.co/how-to/earn-your-leisure-transcripts): download every available transcript from Earn Your Leisure in bulk

## Formats
- [Transcript to Markdown](https://bulktranscripts.co/transcript-to#markdown): Clean headings, metadata, and readable paragraphs, ready for Obsidian, Notion, or any notes app.
- [Transcript to plain text](https://bulktranscripts.co/transcript-to#txt): Nothing but the words. The most portable format there is.
- [Transcript to SRT subtitles](https://bulktranscripts.co/transcript-to#srt): Standard subtitle file with numbered cues and timecodes, ready for any editor.
- [Transcript to WebVTT](https://bulktranscripts.co/transcript-to#vtt): The web-native subtitle format, for HTML5 players and browser captions.
- [Transcript to CSV](https://bulktranscripts.co/transcript-to#csv): One row per caption segment with start time and duration. Open it in a spreadsheet.
- [Transcript to JSON](https://bulktranscripts.co/transcript-to#json): Structured segments with timestamps and full video metadata, for pipelines and code.
- [Transcript to NotebookLM](https://bulktranscripts.co/transcript-to/notebooklm): NotebookLM takes one video at a time. This gives you the whole channel as a single source file you can upload once.
- [Transcript to ChatGPT](https://bulktranscripts.co/transcript-to/chatgpt): A metadata header the model can cite, then clean paragraphs. Paste it in or attach it to a Project.
- [Transcript to Claude](https://bulktranscripts.co/transcript-to/claude): Long context handles an entire channel at once. This is that channel, in one file.

## Company
- [About](https://bulktranscripts.co/about): who builds BulkTranscripts and why
- [Contact](https://bulktranscripts.co/contact): support email and response expectations
- [Changelog](https://bulktranscripts.co/changelog): product and API changes, newest first
- [Privacy](https://bulktranscripts.co/privacy) · [Terms](https://bulktranscripts.co/terms)

## Optional
Legacy web-app API (the /api/v1 endpoints above are preferred):
- POST https://bulktranscripts.co/api/transcript: single video, JSON body {"url": "..."}
- POST https://bulktranscripts.co/api/bulk: playlist/channel, Server-Sent Events stream
- GET  https://bulktranscripts.co/api/export/<video_id>?format=<txt|md|srt|vtt|csv|json|ai>
- POST https://bulktranscripts.co/api/export/bundle: ZIP or combined file for many videos


---

Source: https://bulktranscripts.co/docs
Updated: 2026-10-24

class="">

Developers & no-code builders · 30 free credits

# Get YouTube transcripts into anything

One URL gives you the transcript of any YouTube video as text. Your AI assistant
can call it, your scripts can call it, and tools like n8n or Google Sheets can call it. You
do not need to be a developer to use this page. Pick the path that sounds like you.

### I use an AI assistant

 Claude, ChatGPT, Cursor, VS Code. Connect once, then just ask for transcripts in the chat. No code.

 Connect the MCP server →

### I write code or automate

 Python, Node, curl, n8n, Make, Zapier, Sheets. Create a free key and call one URL.

 Start in 3 steps →

### I just want the files

 Paste a video, playlist, or channel link and download TXT, SRT, Markdown, or JSON.

 Open the web app →

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

 Show
 MCP server
 REST API
 No-code tools
 Everything

## Start in 3 steps

About two minutes. Free: 30 credits, no card.

 - 
 1

### Create a free API key

 Sign in with Google at bulktranscripts.co/app, open the
 MCP & API tab and press Create key. Copy the key. It starts
 with bt_ak_ and is shown only once.

 In plain words: an API key is a password that programs use on your behalf.
 Keep it private, and revoke it from the same place if it ever leaks.

 Open the app and create a key →

 - 
 2

### Ask for one transcript

 Paste this into a terminal (Mac: Terminal; Windows: PowerShell) after replacing
 YOUR_API_KEY. No terminal? Use the playground right
 below instead. It does the same thing with a button.

 curl "https://bulktranscripts.co/api/v1/transcript?video=https://www.youtube.com/watch?v=jNQXAC9IVRw" \
 -H "Authorization: Bearer YOUR_API_KEY"

 In plain words: you are visiting a web address, and the second line shows
 your key so we know whose credits to use.

 - 
 3

### Read the answer

 The reply is JSON: labelled fields, like a form. text is the whole
 transcript, segments are the timestamped lines, and
 billing.remaining is how many credits you have left.

 {
 "video_id": "jNQXAC9IVRw",
 "url": "https://www.youtube.com/watch?v=jNQXAC9IVRw",
 "title": "Me at the zoo",
 "channel": "jawed",
 "duration": 19,
 "duration_text": "0:19",
 "upload_date": "2005-04-23",
 "language": "en",
 "source": "auto_caption",
 "cached": true,
 "word_count": 43,
 "text": "All right, so here we are in front of the elephants...",
 "segments": [{"text": "All right, so here we are...", "start": 0.0, "duration": 5.4}],
 "billing": {"remaining": 29, "creditsCharged": 1, "freeLimit": 30}
}

 Trimmed for reading. The full field list is in the
 reference.

## Try it here, nothing to install

This calls the real API from your browser. Paste your key once (it stays in this browser only),
pick what you want, press Send. The first fetch of a video costs 1 credit;
fetching a video already in your library is free.

 Transcript of a video (1 credit first time)
 Latest uploads of a channel (free)
 Search YouTube (1 credit)

 Send

 No key yet? Create one free, paste it above, press Send.

 // The answer will appear here.

## What you can ask for

Every request is a plain web address (a GET). Replace the parts in UPPERCASE.
Full details for each are in the reference.

I want…Call thisCosts

The transcript of one video/api/v1/transcript?video=VIDEO_URL1 credit, free next timeTry

The transcript as a file (SRT, TXT, Markdown…)/api/v1/transcript?video=VIDEO_URL&format=srtsame as above

Every video in a channel (to fetch them one by one)/api/v1/channel/videos?channel=@HANDLE1 credit for the list

Every video in a playlist/api/v1/playlist/videos?playlist=PLAYLIST_URL1 credit for the list

Videos about a topic/api/v1/search?q=WORDS1 creditTry

A video inside a specific channel/api/v1/channel/search?channel=@HANDLE&q=WORDS1 credit

To know when a channel posts something new/api/v1/channel/latest?channel=@HANDLEfreeTry

To check my balance/api/v1/accountfree

Rule of thumb: getting a transcript costs 1 credit the first time, and reading it
again from your library is free forever. Lists and searches cost 1 credit each. Checking your balance
and watching for new uploads are always free. Anything that fails is not charged.

## Words on this page, in plain English

 APIA way for programs to ask our service for things. You "call" it by visiting a web address.

 EndpointOne specific address that does one job: /api/v1/transcript gets a transcript.

 API keyYour password for programs. It starts with bt_ak_ and links calls to your credits.

 HeaderExtra information sent along with a request. Your key travels in the Authorization header.

 JSONThe format of every answer: labelled fields in curly braces. Every language and no-code tool can read it.

 curlA tiny built-in command-line tool for visiting addresses. Already installed on Mac, Windows, and Linux.

 CreditOne unit of work. 30 free to start; packs never expire.

 MCPThe standard AI assistants use to plug in tools. Connect our server and Claude or ChatGPT can fetch transcripts itself.

 Cache / libraryTranscripts we have already extracted. Yours are free to re-read; everyone's come back instantly.

## Authentication: where the key goes

Every /api/v1/* request, including the free ones, needs a credential. The normal
one is your API key, sent as a Bearer token:

curl "https://bulktranscripts.co/api/v1/transcript?video=VIDEO_ID" \
 -H "Authorization: Bearer bt_ak_YOUR_KEY"

 Pythonheaders={"Authorization": "Bearer bt_ak_…"}

 Node / fetchheaders: { Authorization: "Bearer bt_ak_…" }

 n8n / Make / ZapierHTTP module → Headers → name Authorization, value Bearer bt_ak_…

 Postman / InsomniaAuth tab → Bearer Token → paste the key

 Google SheetsApps Script. See the no-code section

 Any tool with an "API key" fieldHeader name X-API-Key works too

A license key from checkout or an MCP OAuth token (bt_at_…) works
as the Bearer value too. No credential → 401 missing_api_key; wrong one →
401 invalid_api_key. Credits belong to your Google account, so buying a pack tops up
the same balance and the key never changes.

## Credits

Credits are one-time purchases that never expire. No subscription, 30 free
to start.

- 1 credit: a transcript's first addition to your library, a search, or a channel/playlist listing.

- Free: re-reading your library, checking your balance, watching for new uploads, and any failure.

- Every response carries a billing block with creditsCharged and remaining.

- $4.99 · 200
- $14.99 · 1,200
- $29 · 5,000

- $99 · 25,000
- $499 · 200,000

See pricing

## REST API reference

Base URL https://bulktranscripts.co. Every endpoint returns JSON, and all but the bulk-job start
are plain GET requests (machine-readable spec: /openapi.json). Each block shows the same
request in cURL, Python and Node, plus a real response; the table under it lists every option.

 GET/api/v1/transcript
 1 credit first time · your repeats are free

 Full transcript of one YouTube video with metadata, clean paragraph text, and timestamped segments.

 cURLPythonNode / TSResponse
curl "https://bulktranscripts.co/api/v1/transcript?video=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DaircAruvnKk&segments=0" \
 -H "Authorization: Bearer YOUR_API_KEY"
import requests

r = requests.get(
 "https://bulktranscripts.co/api/v1/transcript",
 params={
 "video": "https://www.youtube.com/watch?v=aircAruvnKk",
 "segments": "0"
 },
 headers={"Authorization": "Bearer YOUR_API_KEY"},
)
data = r.json()
print(data["title"], data["word_count"], "words")
const res = await fetch("https://bulktranscripts.co/api/v1/transcript?video=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DaircAruvnKk&segments=0", {
 headers: { Authorization: "Bearer YOUR_API_KEY" },
});
const data = await res.json();
console.log(data.title, data.word_count, "words");
{
 "video_id": "aircAruvnKk",
 "url": "https://www.youtube.com/watch?v=aircAruvnKk",
 "title": "But what is a neural network? | Deep learning chapter 1",
 "channel": "3Blue1Brown",
 "duration": 1120,
 "duration_text": "18:40",
 "upload_date": "2017-10-05",
 "language": "en",
 "source": "manual_caption",
 "cached": true,
 "word_count": 3357,
 "text": "This is a 3. It's sloppily written and rendered at an extremely low resolution...",
 "paragraphs": ["This is a 3. It's sloppily written...", "..."],
 "billing": {"enabled": true, "unlimited": false, "kind": "license", "remaining": 4999,
 "used": 1, "granted": 5000, "creditsCharged": 1}
}

 ParameterTypeDefaultDescription
videostringrequired–YouTube URL, youtu.be link, Shorts URL, or bare 11-char video id.
languagestringoptionalenPreferred caption language(s), comma-separated. Falls back to whatever the video has.
formatenumoptionaljsonjson, or txt, md, srt, vtt, csv, ai to get the rendered file instead of JSON.
segments0 | 1optional10 omits the segment array from JSON (smaller responses).
timestamps0 | 1optional01 includes timestamps in txt/md output.
fresh0 | 1optional01 bypasses the cache and re-extracts. Always charges a credit, even for a video already in your library.

 GET/api/v1/search
 1 credit

 Search YouTube for videos, channels, or playlists. Returns id, title, channel, duration, type, URL and view count per result.

 cURLPythonNode / TSResponse
curl "https://bulktranscripts.co/api/v1/search?q=machine+learning&limit=5" \
 -H "Authorization: Bearer YOUR_API_KEY"
import requests

r = requests.get(
 "https://bulktranscripts.co/api/v1/search",
 params={
 "q": "machine learning",
 "limit": "5"
 },
 headers={"Authorization": "Bearer YOUR_API_KEY"},
)
data = r.json()
for item in data["results"]:
 print(item["id"], item["title"])
const res = await fetch("https://bulktranscripts.co/api/v1/search?q=machine+learning&limit=5", {
 headers: { Authorization: "Bearer YOUR_API_KEY" },
});
const data = await res.json();
for (const item of data.results) console.log(item.id, item.title);
{
 "query": "machine learning",
 "type": "video",
 "results": [
 {
 "id": "aircAruvnKk",
 "url": "https://www.youtube.com/watch?v=aircAruvnKk",
 "title": "But what is a neural network? | Deep learning chapter 1",
 "channel": "3Blue1Brown",
 "channel_url": "https://www.youtube.com/@3blue1brown",
 "duration": 1120,
 "type": "video",
 "view_count": 19234501
 }
 ],
 "count": 5,
 "billing": {"enabled": true, "unlimited": false, "kind": "license", "remaining": 4998, "creditsCharged": 1}
}

 ParameterTypeDefaultDescription
qstringrequired–Search query.
typeenumoptionalvideovideo, channel, or playlist.
limitintegeroptional10Max results, 1–50.

 GET/api/v1/channel/search
 1 credit

 Search within one channel's uploads to find a creator's videos on a topic without listing the whole archive.

 cURLPythonNode / TSResponse
curl "https://bulktranscripts.co/api/v1/channel/search?channel=%40TED&q=AI&limit=10" \
 -H "Authorization: Bearer YOUR_API_KEY"
import requests

r = requests.get(
 "https://bulktranscripts.co/api/v1/channel/search",
 params={
 "channel": "@TED",
 "q": "AI",
 "limit": "10"
 },
 headers={"Authorization": "Bearer YOUR_API_KEY"},
)
data = r.json()
for item in data["results"]:
 print(item["id"], item["title"])
const res = await fetch("https://bulktranscripts.co/api/v1/channel/search?channel=%40TED&q=AI&limit=10", {
 headers: { Authorization: "Bearer YOUR_API_KEY" },
});
const data = await res.json();
for (const item of data.results) console.log(item.id, item.title);
{
 "channel": "@TED",
 "channel_title": "TED",
 "query": "AI",
 "results": [
 {
 "id": "hJP5GqnTrNo",
 "url": "https://www.youtube.com/watch?v=hJP5GqnTrNo",
 "title": "Can AI fix education?",
 "channel": "TED",
 "channel_url": "https://www.youtube.com/@TED",
 "duration": 639,
 "type": "video"
 }
 ],
 "count": 10,
 "billing": {"enabled": true, "unlimited": false, "kind": "license", "remaining": 4997, "creditsCharged": 1}
}

 ParameterTypeDefaultDescription
channelstringrequired–@handle, channel URL, or UC… channel id.
qstringrequired–Topic to search for within the channel.
limitintegeroptional10Max results, 1–50.

 GET/api/v1/channel/videos
 1 credit

 List a channel's videos (up to 1,000) without fetching transcripts. Feed the ids into the transcript endpoint (ids already in your library stay free). There is no pagination: raise limit instead. has_more is true whenever the result reached limit.

 cURLPythonNode / TSResponse
curl "https://bulktranscripts.co/api/v1/channel/videos?channel=%40TED&limit=100" \
 -H "Authorization: Bearer YOUR_API_KEY"
import requests

r = requests.get(
 "https://bulktranscripts.co/api/v1/channel/videos",
 params={
 "channel": "@TED",
 "limit": "100"
 },
 headers={"Authorization": "Bearer YOUR_API_KEY"},
)
data = r.json()
for item in data["results"]:
 print(item["id"], item["title"])
const res = await fetch("https://bulktranscripts.co/api/v1/channel/videos?channel=%40TED&limit=100", {
 headers: { Authorization: "Bearer YOUR_API_KEY" },
});
const data = await res.json();
for (const item of data.results) console.log(item.id, item.title);
{
 "channel": "@TED",
 "title": "TED - Videos",
 "results": [
 {
 "id": "iG9CE55wbtY",
 "url": "https://www.youtube.com/watch?v=iG9CE55wbtY",
 "title": "Do schools kill creativity? | Sir Ken Robinson | TED",
 "channel": "TED",
 "channel_url": "https://www.youtube.com/@TED",
 "duration": 1164
 }
 ],
 "count": 100,
 "has_more": true,
 "billing": {"enabled": true, "unlimited": false, "kind": "license", "remaining": 4996, "creditsCharged": 1}
}

 ParameterTypeDefaultDescription
channelstringrequired–@handle, channel URL, or UC… channel id.
limitintegeroptional100Max videos, 1–1,000.

 GET/api/v1/channel/latest
 free

 A channel's newest uploads (up to 15). Free because it costs us nearly nothing. Poll it for monitoring and only spend credits on what's new. When YouTube's feed is unavailable the response falls back to a direct listing and carries "source": "listing"; published can then be null, so diff on video id, not on date.

 cURLPythonNode / TSResponse
curl "https://bulktranscripts.co/api/v1/channel/latest?channel=%40TED" \
 -H "Authorization: Bearer YOUR_API_KEY"
import requests

r = requests.get(
 "https://bulktranscripts.co/api/v1/channel/latest",
 params={
 "channel": "@TED"
 },
 headers={"Authorization": "Bearer YOUR_API_KEY"},
)
data = r.json()
for item in data["results"]:
 print(item["id"], item["title"])
const res = await fetch("https://bulktranscripts.co/api/v1/channel/latest?channel=%40TED", {
 headers: { Authorization: "Bearer YOUR_API_KEY" },
});
const data = await res.json();
for (const item of data.results) console.log(item.id, item.title);
{
 "channel_id": "UCAuUUnT6oDeKwE6v1NGQxug",
 "channel": "TED",
 "results": [
 {
 "id": "xL2EKQ3Rkos",
 "url": "https://www.youtube.com/watch?v=xL2EKQ3Rkos",
 "title": "A bold plan for sustainable energy",
 "channel": "TED",
 "published": "2026-09-04T14:00:12+00:00",
 "views": "48213"
 }
 ],
 "count": 15,
 "billing_note": "free endpoint — no credit charged"
}

 ParameterTypeDefaultDescription
channelstringrequired–@handle, channel URL, or UC… channel id.

 GET/api/v1/playlist/videos
 1 credit

 Every video in a playlist, in playlist order, so courses stay in sequence. No pagination; has_more is true whenever the result reached limit.

 cURLPythonNode / TSResponse
curl "https://bulktranscripts.co/api/v1/playlist/videos?playlist=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi" \
 -H "Authorization: Bearer YOUR_API_KEY"
import requests

r = requests.get(
 "https://bulktranscripts.co/api/v1/playlist/videos",
 params={
 "playlist": "PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi"
 },
 headers={"Authorization": "Bearer YOUR_API_KEY"},
)
data = r.json()
for item in data["results"]:
 print(item["id"], item["title"])
const res = await fetch("https://bulktranscripts.co/api/v1/playlist/videos?playlist=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi", {
 headers: { Authorization: "Bearer YOUR_API_KEY" },
});
const data = await res.json();
for (const item of data.results) console.log(item.id, item.title);
{
 "playlist": "PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi",
 "title": "Neural networks",
 "results": [
 {
 "id": "aircAruvnKk",
 "url": "https://www.youtube.com/watch?v=aircAruvnKk",
 "title": "But what is a neural network? | Deep learning chapter 1",
 "channel": "3Blue1Brown",
 "channel_url": "https://www.youtube.com/@3blue1brown",
 "duration": 1120
 }
 ],
 "count": 8,
 "has_more": false,
 "billing": {"enabled": true, "unlimited": false, "kind": "license", "remaining": 4995, "creditsCharged": 1}
}

 ParameterTypeDefaultDescription
playliststringrequired–Playlist URL or bare id (the list= value). Ids are 13, 18, 26 or 34 characters; public and unlisted playlists only. A private one reads as nonexistent (playlist_private).
limitintegeroptional100Max videos, 1–1,000.

 GET/api/v1/account
 free

 The credit balance for the key making the call. Returns the same billing block every credited response carries.

 cURLPythonNode / TSResponse
curl "https://bulktranscripts.co/api/v1/account" \
 -H "Authorization: Bearer YOUR_API_KEY"
import requests

r = requests.get(
 "https://bulktranscripts.co/api/v1/account",
 headers={"Authorization": "Bearer YOUR_API_KEY"},
)
data = r.json()
print(data["billing"])
const res = await fetch("https://bulktranscripts.co/api/v1/account", {
 headers: { Authorization: "Bearer YOUR_API_KEY" },
});
const data = await res.json();
console.log(data.billing);
{
 "billing": {"enabled": true, "unlimited": false, "kind": "license", "remaining": 4995,
 "used": 5, "granted": 5000, "freeLimit": 30},
 "docs": "https://bulktranscripts.co/docs"
}

 POST/api/v1/bulk
 listing free · 1 credit per new transcript

 Start a background job that fetches every transcript of a channel, playlist or video URL into your library (up to 1,000 videos). Returns 202 with a run_id straight away; small sources are listed within the call so videos_found is filled, larger ones report it on the first status poll. Cache hits and repeats are free, failures are never charged, and a job that runs out of credits keeps what it fetched and says how many videos it did not attempt. Two jobs per account at a time.

 cURLPythonNode / TSResponse
curl -X POST "https://bulktranscripts.co/api/v1/bulk" \
 -H "Authorization: Bearer YOUR_API_KEY" \
 -H "Content-Type: application/json" \
 -d '{"url": "https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi", "max_videos": 200}'
import requests

r = requests.post(
 "https://bulktranscripts.co/api/v1/bulk",
 json={"url": "https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi", "max_videos": 200},
 headers={"Authorization": "Bearer YOUR_API_KEY"},
)
job = r.json()
print(job["run_id"], job["status"], job["videos_found"])
const res = await fetch("https://bulktranscripts.co/api/v1/bulk", {
 method: "POST",
 headers: { Authorization: "Bearer YOUR_API_KEY", "Content-Type": "application/json" },
 body: JSON.stringify({"url": "https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi", "max_videos": 200}),
});
const job = await res.json();
console.log(job.run_id, job.status, job.videos_found);
{
 "run_id": "3f6f6c0e-6a2b-4b26-9d3e-2f0c1c6f7a11",
 "status": "running",
 "phase": "extracting",
 "source": {"url": "https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi",
 "type": "playlist", "title": "Neural networks"},
 "videos_found": 8,
 "total": 8,
 "completed": 0, "cached": 0, "failed": 0, "skipped": 0, "quota": 0, "interrupted": 0,
 "remaining": 8,
 "error": null,
 "check_again_in_seconds": 15,
 "billing": {"enabled": true, "unlimited": false, "kind": "license", "remaining": 4999,
 "used": 1, "granted": 5000}
}

 ParameterTypeDefaultDescription
urlstringrequired–Channel (@handle or URL), playlist URL or id, or a single video URL. JSON body field.
max_videosintegeroptional100Cap on videos to process, 1 to 1,000. JSON body field.
start_atintegeroptional11-based position to start from, newest first. For sources over 1,000 videos run 1, then 1001, 2001… The status response gives more.next_start_at when a block was full. JSON body field.
contentstringoptionalvideosChannels only: videos (long-form uploads), shorts, or live. JSON body field.
languagestringoptionalenPreferred caption language(s), comma-separated. JSON body field.

 GET/api/v1/bulk/{run_id}
 free

 Progress of a bulk job. status is running, completed, stopped (interrupted by a server restart; start it again, everything already fetched is free) or failed. Wait at least check_again_in_seconds between polls; it is 0 once the job has finished.

 cURLPythonNode / TSResponse
curl "https://bulktranscripts.co/api/v1/bulk/{run_id}" \
 -H "Authorization: Bearer YOUR_API_KEY"
import requests

r = requests.get(
 "https://bulktranscripts.co/api/v1/bulk/{run_id}",
 headers={"Authorization": "Bearer YOUR_API_KEY"},
)
data = r.json()
print(data["status"], data["completed"], "of", data["total"])
const res = await fetch("https://bulktranscripts.co/api/v1/bulk/{run_id}", {
 headers: { Authorization: "Bearer YOUR_API_KEY" },
});
const data = await res.json();
console.log(data.status, data.completed, "of", data.total);
{
 "run_id": "3f6f6c0e-6a2b-4b26-9d3e-2f0c1c6f7a11",
 "status": "completed",
 "phase": "finished",
 "source": {"url": "https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi",
 "type": "playlist", "title": "Neural networks"},
 "videos_found": 8,
 "total": 8,
 "completed": 7, "cached": 2, "failed": 0, "skipped": 1, "quota": 0, "interrupted": 0,
 "remaining": 0,
 "error": null,
 "check_again_in_seconds": 0,
 "billing": {"enabled": true, "unlimited": false, "kind": "license", "remaining": 4994,
 "used": 6, "granted": 5000}
}

 ParameterTypeDefaultDescription
run_idstringrequired–The run_id from the start call, in the path.

 GET/api/v1/bulk/{run_id}/results
 free

 One page of per-video outcomes for a job, in source order: id, title, channel, duration, word count and status (ok, cached, or error with a code). Add include=text to get each finished transcript's text and paragraphs in the same page (25 per page), or include=segments for timestamps too. Every video listed as ok or cached is already in your library, so reading it costs nothing.

 cURLPythonNode / TSResponse
curl "https://bulktranscripts.co/api/v1/bulk/{run_id}/results?limit=2" \
 -H "Authorization: Bearer YOUR_API_KEY"
import requests

r = requests.get(
 "https://bulktranscripts.co/api/v1/bulk/{run_id}/results",
 params={
 "limit": "2"
 },
 headers={"Authorization": "Bearer YOUR_API_KEY"},
)
data = r.json()
for item in data["items"]:
 print(item["video_id"], item["status"], item["title"])
const res = await fetch("https://bulktranscripts.co/api/v1/bulk/{run_id}/results?limit=2", {
 headers: { Authorization: "Bearer YOUR_API_KEY" },
});
const data = await res.json();
for (const item of data.items) console.log(item.video_id, item.status, item.title);
{
 "run_id": "3f6f6c0e-6a2b-4b26-9d3e-2f0c1c6f7a11",
 "status": "completed",
 "total": 8,
 "items": [
 {"position": 0, "video_id": "aircAruvnKk", "title": "But what is a neural network?",
 "status": "ok", "url": "https://www.youtube.com/watch?v=aircAruvnKk",
 "channel": "3Blue1Brown", "duration": 1120, "word_count": 3357},
 {"position": 1, "video_id": "IHZwWFHWa-w", "title": "Gradient descent, how neural networks learn",
 "status": "cached", "url": "https://www.youtube.com/watch?v=IHZwWFHWa-w",
 "channel": "3Blue1Brown", "duration": 1261, "word_count": 3720}
 ],
 "next_cursor": 1,
 "note": "Transcript text is not included here. Every video listed as ok or cached is already in this account's library, so get_transcript (MCP), GET /api/v1/transcript, or this endpoint with include=text returns it without spending a credit."
}

 ParameterTypeDefaultDescription
run_idstringrequired–The run_id from the start call, in the path.
cursorintegeroptional–next_cursor from the previous page; omit for the first page.
limitintegeroptional50Items per page, 1 to 100 (1 to 25 with include).
includestringoptional–text adds text, paragraphs, language and upload date to ok and cached items; segments adds timed segments as well.

## MCP server: YouTube for your AI assistant

Connect https://bulktranscripts.co/mcp once and Claude, ChatGPT, Cursor, VS Code, and 20+ other MCP
clients can fetch transcripts, search YouTube, list channels and playlists, and track new uploads,
inside the conversation. One sign-in (Google or license key) includes 30 free
credits, no card.

Example · once connected, just ask

### Set it up in your client

 Claude
 Claude Code
 Cursor
 ChatGPT
 VS Code
 Other / no OAuth

- Settings → Connectors → Add custom connector

- Name BulkTranscripts, URL https://bulktranscripts.co/mcp → Connect

- On the page that opens, continue with Google (or paste a license key)

claude mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp

Claude Code opens the sign-in on first use (or run /mcp → Authenticate).
Prefer a key? --header "Authorization: Bearer bt_ak_…"

One-click install for Cursor, or add this to ~/.cursor/mcp.json
and sign in under Settings → MCP:

{
 "url": "https://bulktranscripts.co/mcp"
}

- Settings → Connectors → Advanced → Developer mode

- Create a connector: name BulkTranscripts, URL https://bulktranscripts.co/mcp, authentication OAuth

- Sign in on the page that opens

Add to .vscode/mcp.json; VS Code prompts for the sign-in when the server starts:

{
 "mcp": {
 "servers": {
 "bulktranscripts": {
 "type": "http",
 "url": "https://bulktranscripts.co/mcp"
 }
 }
 }
}

Any MCP client takes the universal config:

{
 "url": "https://bulktranscripts.co/mcp"
}

If your client cannot run the OAuth sign-in (n8n, Make, older Codex builds, scripts), create a free
API key at /app → MCP & API and send it as
Authorization: Bearer bt_ak_…, or, when headers are impossible, in the URL:

https://bulktranscripts.co/mcp?key=bt_ak_YOUR_KEY

A key in a URL is stored in that client's config, so treat the whole URL as a secret.

Step-by-step guides:
Claude Code ·
Codex ·
Cursor ·
ChatGPT ·
VS Code ·
n8n ·
Make

### The ten tools

Generated from the same tool schema the server sends to your client, so what
you see here is exactly what Claude, Cursor or ChatGPT sees.

ToolParametersReturnsCost

get_transcriptGet YouTube video transcript
- video string · requiredYouTube video URL or 11-character video id.
- language string · optionalPreferred caption language code, e.g. 'en' or 'de'. Defaults to 'en', falling back to whatever exists.
- include_segments boolean · optionalInclude the timestamped segment list. Default false — the plain text is usually what you want.
- fresh boolean · optionalBypass the cache and re-extract (costs a credit). Default false.
Title, channel, duration, upload date, language, clean paragraph text, word count and (on request) timed segments for one video.1 credit on first library addition · repeat reads free
get_transcriptsGet many transcripts at once
- videos array of string · requiredUp to 20 YouTube video URLs or ids.
- language string · optionalPreferred caption language code. Default 'en'.
Up to 20 transcripts in one call, fetched in parallel; the call returns within about 40 seconds and anything still fetching comes back on the next call, free. Videos without captions are reported per item and do not fail the batch. One call is one request against the rate limit.1 credit per new transcript · repeats free
search_youtubeSearch YouTube
- query string · requiredSearch terms, e.g. 'claude code tutorial'.
- type video | channel | playlist · optionalWhat to search for. Default video.
- limit integer · optionalMax results, default 10.
Videos, channels or playlists matching a query: id, title, channel, duration, URL and view count.1 credit
search_channelSearch inside a channel
- channel string · requiredChannel @handle, URL, or UC… id.
- query string · requiredTopic to search for within the channel.
- limit integer · optionalMax results, default 10.
Videos inside one channel that match a topic, same fields as search.1 credit
get_channel_videosList a channel's videos
- channel string · requiredChannel @handle, URL, or UC… id.
- limit integer · optionalMax videos to list, default 100.
A channel's video list (id, title, duration, URL), up to 1,000, ready to feed into get_transcripts.1 credit
get_playlist_videosList a playlist's videos
- playlist string · requiredPlaylist URL or id (the list= value).
- limit integer · optionalMax videos to list, default 100.
Every video in a playlist, in playlist order.1 credit
get_latest_videosTrack new uploads (free)
- channel string · requiredChannel @handle, URL, or UC… id.
A channel's newest uploads (up to 15) with publish dates and views.Free, always
start_bulk_extractStart a bulk extraction job
- url string · requiredChannel (@handle or URL), playlist URL/id, or a single video URL.
- max_videos integer · optionalCap on videos to process, default 100.
- start_at integer · optional1-based position to start from, newest first (default 1). For sources over 1,000 videos run blocks: start_at=1, then 1001, 2001…; the job status gives next_start_at when a block was full.
- content videos | shorts | live · optionalFor channels: which tab to read. Default 'videos' (long-form uploads); 'shorts' or 'live' for those tabs.
- language string · optionalPreferred caption language code. Default 'en'.
A run_id straight away, plus videos_found once the listing is done (small sources list within the call). The server then fetches every transcript into this account's library in the background.Listing free · 1 credit per new transcript
get_bulk_statusCheck a bulk job
- run_id string · requiredThe run_id of the bulk job.
status, total, completed, failed, skipped, quota, remaining and check_again_in_seconds for a job.Free
list_bulk_resultsList a bulk job's results
- run_id string · requiredThe run_id of the bulk job.
- cursor integer · optionalnext_cursor from the previous page; omit for the first page.
- limit integer · optionalItems per page, default 50.
A page of per-video outcomes (id, title, channel, duration, word count, status, error code) with next_cursor. No transcript text.Free

Which tool the agent picks, prompts that work, troubleshooting

### Which tool does the agent reach for?

You never call these by name. The assistant picks from the descriptions. This is the
mapping it lands on, so you know what a request will cost before you ask.

You say…Tool usedCredits

"Summarize this video: [URL]"get_transcript1 (0 if it is already in your library)

"Compare what these four talks say about X"get_transcripts1 per new video

"Find the best explainers on Y"search_youtube, then get_transcripts for the picks1 + 1 per transcript

"What has @creator said about Z?"search_channel, then get_transcripts1 + 1 per transcript

"Turn this course playlist into study notes"get_playlist_videos, then get_transcripts in batches of 201 + 1 per lecture

"Audit this channel's whole archive"get_channel_videos (up to 1,000), then selective get_transcripts1 + 1 per transcript read

"Did @creator post anything this week?"get_latest_videos0

### Prompts that work on day one

One video

“Summarize this talk with the three strongest quotes and timestamps: https://youtube.com/watch?v=…”

“Pull the transcript and turn it into a LinkedIn post in my voice.”

Several videos

“Here are five interviews with the same founder. Where do their stories contradict each other?”

“Compare how these two channels explain the same concept. Which is clearer for a beginner?”

Research a topic

“Find the most-watched videos about intermittent fasting and tell me where the experts disagree.”

“What has @hubermanlab said about caffeine timing? Cite the episodes.”

Courses & playlists

“Read every lecture in this Stanford playlist and build a study guide with one section per lecture.”

“Which videos in this playlist mention gradient descent?”

Monitoring

“Check @TED, @veritasium and @3blue1brown for new uploads and brief me on anything about AI.”

“Every morning: new videos from these channels, one line each.” (free, uses get_latest_videos)

Whole channels

“List everything @mkbhd has published this year, then read the ten most-viewed and tell me his recurring complaints about phones.”

### Troubleshooting

The connector shows “Needs authentication” or never finishes connecting
Open the sign-in your client offers (Claude Code: /mcp → Authenticate; Claude web:
Settings → Connectors → the connector's menu → Reconnect; Cursor: Settings → MCP → Login). The
sign-in page is served by BulkTranscripts and accepts Google or a license key. If the page never
opens, the client is probably an older build without MCP OAuth: use the key-in-URL form
https://bulktranscripts.co/mcp?key=YOUR_LICENSE_KEY instead.
I get 401 right after buying a pack
The key is issued at checkout but the payment webhook that activates it can land a few seconds
later. Wait a moment and retry; nothing is charged for the failed call.
The assistant answers from memory instead of calling the tool
Paste the full YouTube URL (not just the title) and say “use BulkTranscripts” or “fetch the
transcript” once. Check the connector is enabled for the current chat; ChatGPT and Claude web both
have a per-conversation toggle.
out_of_credits in the middle of a batch
The transcripts already fetched are returned with partial: true and a
stopped count, and the error carries a purchase link. Credits are added to the same
account you signed in with, so nothing needs reconnecting afterwards.
no_transcript for a video that clearly has captions
Some videos only have captions in one language; ask for that language explicitly
(“get the German transcript”). Live streams and members-only videos expose no captions. Never
charged.
Long videos (2h+) take a while or the client times out
Extraction usually finishes in a few seconds, but the first read of a long, uncached video can
take longer than a strict client timeout. Ask again: the second request returns instantly from the
cache and the credit is not charged twice.
Something else
Every error has a stable code listed in the error table.
Still stuck? Email hello@bulktranscripts.co with the tool name and the code.

## Agent skill (Claude Code, Codex, OpenClaw & friends)

Prefer a skill over a server? One command installs a skill that teaches your agent the whole
API. Create a free API key first at /app → MCP & API:

mkdir -p ~/.claude/skills/youtube-transcripts
curl -fsSL https://bulktranscripts.co/skill.md -o ~/.claude/skills/youtube-transcripts/SKILL.md

OpenClaw users install it from ClawHub instead: openclaw skills install
@pratie/bulktranscripts-youtube (listing:
clawhub.ai/pratie/skills/bulktranscripts-youtube).

Set BULKTRANSCRIPTS_API_KEY in your environment to your
bt_ak_… key (a license key works too). The skill source lives at /skill.md and on
GitHub;
capabilities, example prompts, and per-agent notes are on the
agent skill page.

## No-code tools: n8n, Make, Zapier, Google Sheets

Anything with an "HTTP request" step can use the API. Method GET, the URL from
the table above, and one header: Authorization =
Bearer bt_ak_…. Then read text from the JSON reply.

- n8n: HTTP Request node. Step-by-step: n8n guide.

- Make: HTTP → Make a request. Step-by-step: Make guide.

- Zapier: Webhooks by Zapier → GET, add the header under "Headers".

- Google Sheets: Extensions → Apps Script, paste this, then use
=TRANSCRIPT(A2) in a cell:

const KEY = "bt_ak_YOUR_KEY";
function TRANSCRIPT(video) {
 const url = "https://bulktranscripts.co/api/v1/transcript?segments=0&video=" + encodeURIComponent(video);
 const res = UrlFetchApp.fetch(url, {headers: {Authorization: "Bearer " + KEY}, muteHttpExceptions: true});
 const data = JSON.parse(res.getContentText());
 return data.text || (data.error && data.error.message) || "";
}

Sheets recalculates formulas on every edit; each recalculation of a video
already in your library is free, so the cost is one credit per new video.

## Questions people ask

Do I need to know how to code?
No. The web app needs
nothing, the MCP server is a one-time connect inside your AI assistant, and
the no-code tools only need a URL and a header. The code samples are
for people who want them.

Is it really free to start?
Yes. 30 credits with a Google
sign-in, no card. That is 30 transcripts, searches or listings. Re-reading a transcript you
already fetched is free forever.

Where do I find my API key later?
You can't. It is shown once when
created. Open MCP & API, revoke the old one, create a new one. Keys are
free and you can have five.

What happens when I run out?
The request returns
402 out_of_credits with a purchase link. Buy a pack while signed in with the same
Google account and your existing key keeps working, with nothing to reconnect.

Can I get a whole channel at once?
Yes, up to 1,000 videos per job.
Over the API: POST /api/v1/bulk with the channel or playlist URL, then poll
/api/v1/bulk/{run_id}. Over MCP: start_bulk_extract, then
get_bulk_status. The server keeps fetching after the call returns; each new transcript
costs 1 credit and lands in your library. With zero code, the web app does the
whole channel and zips the files.

Why did a video come back with "no_transcript"?
It has no captions at
all, not even auto-generated ones. You are not charged. Try another language first;
if that fails, the video simply has none.

Is my key safe in the playground?
It is stored only in your own
browser's local storage, sent only to bulktranscripts.co, and never logged. Clear the field to
forget it.

## Errors

Every error is JSON with a stable code and a message that says what to do next:

{
 "error": {"code": "no_transcript", "message": "No captions available for this video..."}
}

HTTPCodeMeaningRetry?

400invalid_inputMissing or malformed video / channel / playlist / q parameter, or an unknown format. A caller bug.No. Fix the request.

400resolution_failedA well-formed id or URL that could not be resolved (private, removed, region-locked, nonexistent). Not charged.No.

400empty_sourceThe channel or playlist resolved but contains no videos.No.

400playlist_privateYouTube reports the playlist as nonexistent, which is also what it says about a private playlist. Ask the owner to set Visibility to Unlisted or Public, then retry. Not charged.After the visibility change.

400private_video / members_only / age_restrictedYouTube will not serve this video's captions without a viewer login. Not charged; in a bulk job it counts as skipped.No.

401missing_api_keyNo credential on the request. Sign in at /app and create a key. The error object carries setup_url and docs_url.No. Add the key.

401invalid_api_keyThe Bearer credential is not a valid API key or license key.No. Check the key.

401invalid_tokenThe OAuth access token is expired or revoked. Refresh or reconnect.After refreshing or reconnecting.

402out_of_creditsBalance empty. The error object also carries purchase_url, setup_url, and reconnect_steps so an agent can guide the user.After topping up.

403access_blockedThe caller has been blocked for abuse.No.

404no_transcriptThe video has no captions at all. Not charged.No. Try another language first.

404not_foundUnknown /api/v1 path.No.

429rate_limitedToo many requests. Wait the number of seconds in the Retry-After header, then retry.Yes, after Retry-After seconds.

429too_many_runsThis account already has 2 bulk jobs running. Let one finish, then start the next.Yes, after Retry-After seconds.

500internal_errorSomething broke on our side. Retry once, then report it.Yes, once.

503restartingThe server is restarting for an update; back within a minute.Yes, in a minute.

504still_fetchingThe transcript is still being fetched and took longer than the request allows. It keeps going in the background.Yes. The same request returns it, usually within a minute.

Over MCP the same codes arrive as a tool result with isError: true
and the identical error.code. A get_transcripts batch that runs out of
credits part-way returns the transcripts it completed plus partial: true and a
stopped count, so nothing already paid for is lost.

## Rate limits

Per public IP, shared between REST and MCP traffic from that IP:
120 requests/minute overall, and 30 requests/minute for
anything under /api/v1/ (including /account and cache hits), plus
any MCP tool call. Exceeding either returns 429 rate_limited with a
Retry-After header in seconds. Bulk jobs have their own budget: 12 starts per
10 minutes per IP, and 2 running at once per account. The REST API fetches one transcript per request;
for batches, pace at under 30 per minute or use the MCP get_transcripts tool,
which fetches up to 20 videos in parallel and counts as a single request. If you are building something high-volume (100k+ transcripts/month),
email us for a custom pack.

Stuck anywhere on this page? Email
hello@bulktranscripts.co with what you tried
and what you see. A person answers, usually within a day.

 On this page

 Start in 3 steps
 Try it here
 What you can ask for
 Plain-English glossary
 Authentication
 REST API reference
 Agent skill
 MCP server
 The ten tools
 No-code tools
 Credits
 Questions
 Errors
 Rate limits
 OpenAPI spec ↗
 llms.txt ↗
 Integration guides ↗

Updated October 24, 2026

---

Source: https://bulktranscripts.co/youtube-mcp-server
Updated: 2026-10-24

class="">

 Hosted YouTube MCP server · 30 free credits
 
# YouTube MCP server: give ChatGPT and Claude a whole channel

 One prompt pulls the transcripts of up to 1,000 videos: a channel,
 a playlist, a podcast's back catalogue. Your assistant then reads what was actually said, not
 just the titles. It can also search YouTube and watch channels for new uploads. Nothing to install.

 Server URLhttps://bulktranscripts.co/mcp

 Add it as a connector in your app, sign in with Google, then just ask.
 Steps for your app ↓

 Works with
 GChatGPTClaudeClaude CodeCursorVS CodeCCodexXGrok

 ChatGPT · BulkTranscripts connected

 Find the 15 videos on this channel that explain how they build AI agents, and turn them into one playbook.

 search_channel15 relevant videos1 credit

 get_transcripts15 full transcripts15 credits

 Your playbook, built from all 15 transcripts
 - Every video and its core idea, in one table

 - Rules, workflows, architecture and prompts in ten sections

 - Each point traced to the video it came from

 The real flow from our 2-minute ChatGPT recording

 1,000videos per bulk joba whole channel or playlist

 10tools for your assistanttranscripts, search, uploads

 30free credits to startGoogle sign-in, no card

 $0to read a video againyour library stays free

## What is a YouTube MCP server?
An MCP (Model Context Protocol) server gives an AI assistant such as Claude, ChatGPT or Cursor a set of tools it can call. The BulkTranscripts MCP server is hosted, so there is nothing to run: connect it with one URL and one Google sign-in and the assistant can fetch transcripts, search YouTube, list channels and playlists, and watch for new uploads.

00:30Setup
## Connect it in about a minute
Pick your app. Every one ends the same way: one Google sign-in, and the account starts with 30 free credits.

ClaudeChatGPTClaude CodeCursorVS CodeCodexOther / API key

### Claude (web & desktop): add a custom connector
In Claude, open Settings → Connectors → Add custom connector. Name it BulkTranscripts, paste the URL below, press Add, then Connect. A sign-in page opens: continue with Google and the account includes 30 free credits. Then just ask, for example "Get the transcript of this video and summarize it" with a YouTube link.
https://bulktranscripts.co/mcp

### ChatGPT: custom connector
Settings → Connectors → Advanced → enable Developer mode, then Create a connector named BulkTranscripts with the URL below and OAuth authentication. Sign in with Google when prompted. Step by step on the ChatGPT guide.
https://bulktranscripts.co/mcp

### Claude Code: one command
Add the server, then run /mcp in a session and pick Authenticate. A sign-in page opens; continue with Google and the account includes 30 free credits. Full walkthrough: Claude Code guide.
claude mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp

### Cursor: one JSON block
Put this in ~/.cursor/mcp.json (global) or .cursor/mcp.json in a repo. Under Settings → MCP the server shows a login prompt; sign in once. There is also a one-click install link on the Cursor guide.
{
 "mcpServers": {
 "bulktranscripts": { "url": "https://bulktranscripts.co/mcp" }
 }
}

### VS Code Copilot: .vscode/mcp.json
VS Code 1.101+ supports remote HTTP servers and MCP auth natively. Create the file, reload, and confirm the tools appear under Copilot Chat's Agent mode. Details on the VS Code guide.
{
 "servers": {
 "bulktranscripts": { "type": "http", "url": "https://bulktranscripts.co/mcp" }
 }
}

### Codex CLI: config.toml
Recent Codex builds speak streamable-HTTP MCP. Builds with MCP OAuth open the sign-in on first use; on builds without it, use the key-in-URL form shown on the Codex guide.
[mcp_servers.bulktranscripts]
url = "https://bulktranscripts.co/mcp"

### Skip OAuth with a license key
Any client that can send headers can attach a key directly. Clients that cannot (ChatGPT's connector form, some Codex builds) accept the key in the URL instead: https://bulktranscripts.co/mcp?key=YOUR_KEY. Nothing to buy: create a free API key at /app → MCP & API → API keys and it works here and on the REST API. A license key from a credit pack works in the same slot.
claude mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp \
 --header "Authorization: Bearer YOUR_LICENSE_KEY"

01:10Ten tools
## What your assistant can do with it
Ten tools in five jobs. Costs are per call; anything already in your library is free to read again.

### Read transcripts
The full text with title, channel and timestamps, so answers quote what was actually said.
- get_transcript1 video1 credit
- get_transcriptsup to 20 at once1 per video

### Pull a whole channel
A channel, playlist or podcast back catalogue in one job. Runs on the server, so the chat never times out.
- start_bulk_extractup to 1,000 videos1 per video
- get_bulk_statusprogressfree
- list_bulk_resultswhat landedfree

### Find the right videos
Search all of YouTube, or search inside one creator's channel for a topic.
- search_youtubeby keyword1 credit
- search_channelinside a channel1 credit

### List an archive
Every video on a channel or in a playlist, so the assistant picks what to read.
- get_channel_videosup to 1,0001 credit
- get_playlist_videosin order1 credit

### Watch for new uploads
“Did anyone I follow post this week?” Poll it every morning; it never costs a thing.
- get_latest_videosnewest uploadsfree

### You just ask
You never type a tool name. Ask in plain words and the assistant picks the tool, the way it picks a search engine.
Which tool runs for which request ↓

01:52Watch
## One prompt, 15 transcripts, one playbook
A real recording in ChatGPT, sped up where it waits.

Watch: one ChatGPT prompt, 15 transcripts, one playbook (2:22, narrated). 

02:24Ideas
## Things people ask it
Paste a link or name a channel. The bigger the archive, the more it saves you.

### A whole podcast
“Pull every episode of this podcast and list the books the host keeps recommending, with the episodes that mention them.”

### A course playlist
“Read all the lectures in this playlist and build a study guide with one section per lecture.”

### Competitor channels
“What have these three channels posted this month? One line each, then the themes they share.”

### One creator, one topic
“What has @hubermanlab said about caffeine timing? Quote the episodes.”

### Where experts disagree
“Find the most-watched videos on intermittent fasting and tell me where the experts disagree.”

### Every morning
“Check @TED, @veritasium and @3blue1brown for new uploads and brief me on anything about AI.” free

03:05Hosted
## Why a hosted server

### Nothing to run
No npx, Docker or keys to juggle. BulkTranscripts runs the extraction (proxy routing, retries, a shared cache) and keeps it working when YouTube changes.

### Pay once, re-read free
30 free credits with Google sign-in. A transcript costs 1 credit the first time it enters your library; reading it again is free in every app. One-time packs that never expire, no subscription.

### Built for whole channels
Bulk jobs run up to 1,000 videos on the server and report progress, so long jobs survive the chat. Videos without captions are skipped and never charged.

## Hosted server vs. the GitHub MCP servers

BulkTranscripts operates the extraction service for you. Self-hosting gives you control of
the implementation and deployment, with responsibility for its maintenance. Features vary by
repository: check the specific server's documentation before choosing. This is a deployment
comparison written by BulkTranscripts, not a benchmark of every alternative.

BulkTranscripts MCPSelf-hosted repos (npx / uvx)

SetupHosted URL with OAuth or a license keyRuntime and hosting depend on the chosen repository

Where requests come fromManaged extraction with residential routing and retries; failures remain possibleThe network or proxy you configure

ToolsTranscript, batch of 20, YouTube search, channel search, channel list, playlist list, free new-upload checkCheck the repository's tool list

Whole channel or playlistList up to 1,000 videos, then batch-fetchCheck listing, pagination and batch support

CacheShared cache; repeat reads from your library are freeDepends on implementation and configuration

AuthenticationOAuth sign-in or a license key header; credits follow the accountDepends on implementation and deployment

Cost30 free, then one-time packs from $4.99 for 200; credits never expireCheck licensing, hosting, network and maintenance costs

 03:40Reference
## Which tool runs, and what it costs

You never call these by name. The assistant picks from the descriptions. This is the
mapping it lands on, so you know what a request will cost before you ask.

You say…Tool usedCredits

"Summarize this video: [URL]"get_transcript1 (0 if it is already in your library)

"Compare what these four talks say about X"get_transcripts1 per new video

"Find the best explainers on Y"search_youtube, then get_transcripts for the picks1 + 1 per transcript

"What has @creator said about Z?"search_channel, then get_transcripts1 + 1 per transcript

"Turn this course playlist into study notes"get_playlist_videos, then get_transcripts in batches of 201 + 1 per lecture

"Audit this channel's whole archive"get_channel_videos (up to 1,000), then selective get_transcripts1 + 1 per transcript read

"Did @creator post anything this week?"get_latest_videos0

 Full parameters and return fields for every tool

Generated from the same tool schema the server sends to your client, so what
you see here is exactly what Claude, Cursor or ChatGPT sees.

ToolParametersReturnsCost

get_transcriptGet YouTube video transcript
- video string · requiredYouTube video URL or 11-character video id.
- language string · optionalPreferred caption language code, e.g. 'en' or 'de'. Defaults to 'en', falling back to whatever exists.
- include_segments boolean · optionalInclude the timestamped segment list. Default false — the plain text is usually what you want.
- fresh boolean · optionalBypass the cache and re-extract (costs a credit). Default false.
Title, channel, duration, upload date, language, clean paragraph text, word count and (on request) timed segments for one video.1 credit on first library addition · repeat reads free
get_transcriptsGet many transcripts at once
- videos array of string · requiredUp to 20 YouTube video URLs or ids.
- language string · optionalPreferred caption language code. Default 'en'.
Up to 20 transcripts in one call, fetched in parallel; the call returns within about 40 seconds and anything still fetching comes back on the next call, free. Videos without captions are reported per item and do not fail the batch. One call is one request against the rate limit.1 credit per new transcript · repeats free
search_youtubeSearch YouTube
- query string · requiredSearch terms, e.g. 'claude code tutorial'.
- type video | channel | playlist · optionalWhat to search for. Default video.
- limit integer · optionalMax results, default 10.
Videos, channels or playlists matching a query: id, title, channel, duration, URL and view count.1 credit
search_channelSearch inside a channel
- channel string · requiredChannel @handle, URL, or UC… id.
- query string · requiredTopic to search for within the channel.
- limit integer · optionalMax results, default 10.
Videos inside one channel that match a topic, same fields as search.1 credit
get_channel_videosList a channel's videos
- channel string · requiredChannel @handle, URL, or UC… id.
- limit integer · optionalMax videos to list, default 100.
A channel's video list (id, title, duration, URL), up to 1,000, ready to feed into get_transcripts.1 credit
get_playlist_videosList a playlist's videos
- playlist string · requiredPlaylist URL or id (the list= value).
- limit integer · optionalMax videos to list, default 100.
Every video in a playlist, in playlist order.1 credit
get_latest_videosTrack new uploads (free)
- channel string · requiredChannel @handle, URL, or UC… id.
A channel's newest uploads (up to 15) with publish dates and views.Free, always
start_bulk_extractStart a bulk extraction job
- url string · requiredChannel (@handle or URL), playlist URL/id, or a single video URL.
- max_videos integer · optionalCap on videos to process, default 100.
- start_at integer · optional1-based position to start from, newest first (default 1). For sources over 1,000 videos run blocks: start_at=1, then 1001, 2001…; the job status gives next_start_at when a block was full.
- content videos | shorts | live · optionalFor channels: which tab to read. Default 'videos' (long-form uploads); 'shorts' or 'live' for those tabs.
- language string · optionalPreferred caption language code. Default 'en'.
A run_id straight away, plus videos_found once the listing is done (small sources list within the call). The server then fetches every transcript into this account's library in the background.Listing free · 1 credit per new transcript
get_bulk_statusCheck a bulk job
- run_id string · requiredThe run_id of the bulk job.
status, total, completed, failed, skipped, quota, remaining and check_again_in_seconds for a job.Free
list_bulk_resultsList a bulk job's results
- run_id string · requiredThe run_id of the bulk job.
- cursor integer · optionalnext_cursor from the previous page; omit for the first page.
- limit integer · optionalItems per page, default 50.
A page of per-video outcomes (id, title, channel, duration, word count, status, error code) with next_cursor. No transcript text.Free

 Troubleshooting

The connector shows “Needs authentication” or never finishes connecting
Open the sign-in your client offers (Claude Code: /mcp → Authenticate; Claude web:
Settings → Connectors → the connector's menu → Reconnect; Cursor: Settings → MCP → Login). The
sign-in page is served by BulkTranscripts and accepts Google or a license key. If the page never
opens, the client is probably an older build without MCP OAuth: use the key-in-URL form
https://bulktranscripts.co/mcp?key=YOUR_LICENSE_KEY instead.
I get 401 right after buying a pack
The key is issued at checkout but the payment webhook that activates it can land a few seconds
later. Wait a moment and retry; nothing is charged for the failed call.
The assistant answers from memory instead of calling the tool
Paste the full YouTube URL (not just the title) and say “use BulkTranscripts” or “fetch the
transcript” once. Check the connector is enabled for the current chat; ChatGPT and Claude web both
have a per-conversation toggle.
out_of_credits in the middle of a batch
The transcripts already fetched are returned with partial: true and a
stopped count, and the error carries a purchase link. Credits are added to the same
account you signed in with, so nothing needs reconnecting afterwards.
no_transcript for a video that clearly has captions
Some videos only have captions in one language; ask for that language explicitly
(“get the German transcript”). Live streams and members-only videos expose no captions. Never
charged.
Long videos (2h+) take a while or the client times out
Extraction usually finishes in a few seconds, but the first read of a long, uncached video can
take longer than a strict client timeout. Ask again: the second request returns instantly from the
cache and the credit is not charged twice.
Something else
Every error has a stable code listed in the error table.
Still stuck? Email hello@bulktranscripts.co with the tool name and the code.

 Building against the REST API instead? Same engine, same credits:
 API docs · OpenAPI spec ·
 agent skill.

04:15Questions
## Questions

Which MCP clients does it work with?
Anything that speaks MCP over streamable HTTP: Claude Code, Claude web/desktop connectors, ChatGPT developer mode, Cursor, Windsurf, VS Code, Codex, Cline, Grok, and the official MCP SDKs.

Is it really free to try?
Yes. Signing in with Google includes 30 free transcript additions, no card needed. The latest-uploads tool is free forever.

Do I have to run anything locally?
No. It is a hosted remote server at https://bulktranscripts.co/mcp: no npx, no uvx, no Docker, no API key juggling to start.

How is this different from the open-source YouTube transcript MCP servers?
BulkTranscripts is hosted and includes search, channel and playlist listing, batch fetching, a shared cache, and account-based authentication. Self-hosted servers vary in features and give you control of deployment and maintenance. Check each repository's tools, limits, authentication, and network requirements.

Can it read a whole channel?
Yes: the agent lists the channel's videos (up to 1,000), then batch-fetches transcripts. Videos already in your account's library come back instantly and free.

## Hand your assistant its first channel

 Copy the URL, add it as a connector, sign in with Google. 30 free credits, no card.

 Server URLhttps://bulktranscripts.co/mcp

 Stuck on setup? Email hello@bulktranscripts.co with the app or code you are using and what you see. A person answers, usually within a day.

## More YouTube transcript tools

Updated October 24, 2026

---

Source: https://bulktranscripts.co/youtube-transcript-api
Updated: 2026-10-24

class="">

 Developer API · 30 free credits
 
# YouTube transcript API

 Extract one video or build a bulk transcript pipeline for thousands. Get clean JSON with timestamps, search, channel and playlist endpoints, plus a hosted YouTube MCP server for AI agents.

 Read the API docs
 
 Works withGChatGPTClaudeClaude CodeCursorCCodexn8nMake

 Your first call, right now
curl "https://bulktranscripts.co/api/v1/transcript?video=https://www.youtube.com/watch?v=jNQXAC9IVRw" \
 -H "Authorization: Bearer YOUR_API_KEY"

30 free credits with a free API key. Repeat reads are free.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

## What is the BulkTranscripts YouTube transcript API?
A hosted REST API that returns clean transcript JSON for any YouTube video, plus endpoints to search YouTube, list a channel's or playlist's videos, and poll a channel's latest uploads for free. It runs on residential proxies so it is not blocked the way self-hosted scrapers are, and repeat reads of a transcript already in your own library are free. Every call needs an API key; creating one is free and includes 30 credits.

00:40Developer API
## How it works
Everything below runs on the same engine as the web app.

### Which YouTube transcript workflow does this API fit?
BulkTranscripts is a hosted YouTube transcript API for developers and researchers who need more than a one-video downloader. Use the transcript endpoint for a list of YouTube URLs, the channel endpoint to enumerate up to 1,000 uploads, or the playlist endpoint to preserve a course or series in order. Each successful extraction returns clean text, video metadata, language, word count, and timestamped segments that can be stored as JSON or rendered as TXT, Markdown, SRT, VTT, CSV, or an AI-ready file. Search and latest-upload endpoints cover discovery and monitoring, while your own library makes repeated reads and re-exports free. For Claude, ChatGPT, Codex, Cursor, VS Code, and other MCP-capable assistants, the same capabilities are available through the hosted YouTube MCP server; automation tools such as n8n and Make call the REST API directly. Choose the REST API when your application controls batching and storage; choose MCP when an AI assistant should search YouTube, list channel videos, and retrieve transcripts as tools inside a conversation.

### Your first call, right now
One GET request, one free API key, no SDK. Create the key at /app → MCP & API → API keys (Google sign-in, 30 credits, no card). You get title, channel, duration, upload date, language, the full text, clean paragraphs, and a segments array of {text, start, duration} objects.
curl "https://bulktranscripts.co/api/v1/transcript?video=https://www.youtube.com/watch?v=jNQXAC9IVRw" \
 -H "Authorization: Bearer YOUR_API_KEY"

### What you get back
Clean JSON you can use from any language. Need a file instead? Add format=srt (or txt, md, vtt, csv, ai) and the same endpoint returns the rendered document.
{
 "title": "Me at the zoo",
 "channel": "jawed",
 "duration": 19,
 "language": "en",
 "cached": true,
 "word_count": 43,
 "text": "All right, so here we are in front of the elephants...",
 "segments": [{"text": "All right...", "start": 0.0, "duration": 5.4}],
 "billing": {"remaining": 29, "creditsCharged": 1}
}

### More than transcripts
/api/v1/search finds videos by keyword. /api/v1/channel/videos lists up to 1,000 videos from any channel. /api/v1/playlist/videos keeps playlist order, so courses stay in sequence. /api/v1/channel/latest returns a channel's newest uploads for free, so you can poll it to trigger pipelines and only spend credits on new videos.

### A shared cache makes it fast and cheap
Every transcript anyone extracts lands in a shared cache, so cache hits return in milliseconds without another network fetch. The first addition to your library costs one credit; repeat lookups and channel re-runs from your own library are free. Failed extractions are refunded automatically.

### No YouTube blocks to fight
Running youtube-transcript-api or yt-dlp from a datacenter IP means bot checks, empty responses, and constant proxy babysitting. That routing is our problem: residential proxies, proof-of-origin tokens, and retries are built in, so the same request that works on your laptop works in production.

### Pricing that respects side projects
Credits are one-time purchases, not a subscription: $4.99 for 200, $14.99 for 1,200, $29 for 5,000, $99 for 25,000, $499 for 200,000. They never expire, and credits top up the account your API key already belongs to, so nothing to re-paste. (The license key you get at checkout works as an API key too.) Signing in gets you 30 free transcripts, no card required.

02:10Questions
## Questions

Do I need an API key?
Yes, and it is free. Sign in with Google at /app, open MCP & API → API keys, and create one: it starts with <code>bt_ak_</code> and comes with 30 free credits, no card. Credits follow the account, so buying a pack later keeps the same key working. A license key from checkout also works as an API key.

How is this different from the youtube-transcript-api Python library?
The library runs on your IP, so it gets blocked the moment you deploy to a cloud host. This is a hosted service with residential routing, retries, a shared cache, search and channel endpoints, and JSON you can use from any language.

Does it return timestamps?
Yes. Every segment carries start and duration in seconds, and SRT/VTT output is timestamped by definition.

Is there an OpenAPI spec?
Yes, at /openapi.json. Generate a typed client in any language, or hand it to an AI agent.

## Start with 30 free credits
Sign in with Google, create a key, make the first call. No card.
Read the API docs
Stuck on setup? Email hello@bulktranscripts.co with the app or code you are using and what you see. A person answers, usually within a day.

## More YouTube transcript tools

Updated October 24, 2026

---

Source: https://bulktranscripts.co/youtube-transcript-agent-skill
Updated: 2026-10-24

class="">

 Agent skill · 30 free credits
 
# Give your AI agent access to YouTube transcripts

 One file, installed once, teaches your agent the whole YouTube transcript REST API. No MCP server to configure, just a free API key from the app.

 Install the skill
 
 Works withGChatGPTClaudeClaude CodeCursorCCodexn8nMake

 Install in one command
mkdir -p ~/.claude/skills/youtube-transcripts
curl -fsSL https://bulktranscripts.co/skill.md \
 -o ~/.claude/skills/youtube-transcripts/SKILL.md

30 free credits with a free API key. Repeat reads are free.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

## What is the YouTube transcript agent skill?
An agent skill is a Markdown file (SKILL.md) that teaches a coding agent such as Claude Code or Codex how to use a service. The BulkTranscripts skill documents the whole REST API, so the agent can fetch YouTube transcripts, search channels and monitor uploads with plain HTTP calls, no MCP client required. Install it by saving one file, then give the agent a free API key created at /app under MCP & API.

00:40Agent skill
## How it works
Everything below runs on the same engine as the web app.

### Install in one command
Works for Claude Code today, and for Codex use ~/.codex/skills/ instead. Any agent that reads the standard SKILL.md format works the same way. The raw file lives at /skill.md if you want to read it first (you should, since it is short), and the source, examples, and issues live on GitHub: pratie/youtube-transcript-skill.
mkdir -p ~/.claude/skills/youtube-transcripts
curl -fsSL https://bulktranscripts.co/skill.md \
 -o ~/.claude/skills/youtube-transcripts/SKILL.md

### OpenClaw: install from ClawHub
The same skill is published on ClawHub, OpenClaw's skill registry, as @pratie/bulktranscripts-youtube (security audit: pass). One command installs it into an OpenClaw workspace; npx clawhub install bulktranscripts-youtube does the same for registry-managed skill folders.
openclaw skills install @pratie/bulktranscripts-youtube

### What your agent can do with it
Fetch any video's transcript as clean text with timestamps and metadata. Search YouTube for videos, channels, or playlists. Search inside one channel's uploads. List a channel's videos (up to 1,000) or a playlist in order. Track a channel's newest uploads for free. Every capability is a documented curl call, so it works in any agent with shell access.

### Why a skill instead of an MCP server
Skills are plain instructions: no process to run, no client-specific config, no protocol version to negotiate. Agent Skills are now a standardized, portable format across Claude Code, Codex, OpenClaw, and a growing list of agents: install once, carried in a folder, versioned in git if you like. If you prefer MCP, the same engine is a hosted server. See the YouTube MCP server page.

### One key, set once
Sign in with Google at /app, open MCP & API → API keys, and create a key: free, 30 transcript credits, no card. Put it in BULKTRANSCRIPTS_API_KEY and the skill sends it as a Bearer token on every call. Buying a one-time credit pack tops up the same account, so the key never changes. Credits never expire, repeat reads from your own library are free, and the skill checks the balance in every response.

### Example prompts that just work
"Summarize this video and pull the key quotes." · "What has @TED published about AI? Fetch the three most relevant transcripts." · "Check if this channel posted anything this week and brief me." · "Get every transcript in this course playlist and turn it into study notes." The skill includes these playbooks, so the agent knows to list before fetching and to use the free latest-uploads endpoint for monitoring.

02:10Questions
## Questions

Which agents support this skill?
Anything that reads the standard SKILL.md format: Claude Code, OpenAI Codex, OpenClaw (via ClawHub), and compatible agents. Plus, any agent with shell access can simply follow the documented curl calls.

Is it free?
The skill file is free and redistributable. Usage rides the same allowance as everything else: 30 free transcript extractions, then one-time credit packs from $4.99. The new-uploads tracker is free forever.

How is this different from the MCP server?
Same engine, same credits, different transport. MCP suits clients with built-in server support (Claude connectors, Cursor, ChatGPT). The skill suits terminal agents and anywhere you'd rather have zero configuration.

Can I inspect or modify the skill?
Yes. It is a single markdown file at bulktranscripts.co/skill.md. Save it, edit the playbooks, commit it to your dotfiles. The skill file is freely redistributable; the API it calls is a hosted, proprietary service.

## Start with 30 free credits
Sign in with Google, create a key, make the first call. No card.
Install the skill
Stuck on setup? Email hello@bulktranscripts.co with the app or code you are using and what you see. A person answers, usually within a day.

## More YouTube transcript tools

Updated October 24, 2026

---

Source: https://bulktranscripts.co/youtube-api-for-ai-agents
Updated: 2026-10-24

class="">

 For AI builders · 30 free credits
 
# YouTube transcripts for AI agents

 Whatever your agent stack (raw HTTP, MCP, or agent skills), the same engine feeds it clean, cited, timestamped YouTube text.

 See the integration docs
 
 Works withGChatGPTClaudeClaude CodeCursorCCodexn8nMake

 Three ways in, one engine
REST APIcurl "https://bulktranscripts.co/api/v1/transcript?video=VIDEO_ID" \
 -H "Authorization: Bearer YOUR_API_KEY"

MCP serverclaude mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp

Agent skillcurl -fsSL https://bulktranscripts.co/skill.md \
 -o ~/.claude/skills/youtube-transcripts/SKILL.md

Same credits and the same library behind each one.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

## How can an AI agent get YouTube transcripts?
Three ways, all backed by the same engine: the BulkTranscripts REST API (any language, a free API key as Bearer token), the hosted MCP server (Claude, ChatGPT, Cursor, VS Code, Codex), or the agent skill file for coding agents. Each returns clean transcript JSON, supports search and channel listing, and offers free new-upload polling for monitoring.

00:40For AI builders
## How it works
Everything below runs on the same engine as the web app.

### Three integration surfaces, one engine
A REST API for code and no-code tools (n8n, Make, Zapier call it with one HTTP node). An MCP server for Claude, ChatGPT, Cursor, and friends. And an agent skill: a single SKILL.md your coding agent installs once and then knows the whole API. All three share credits, cache, and the same underlying extraction chain; per-client setup lives in the integration guides.

### Built for RAG and pipelines
Transcript responses carry everything a retrieval pipeline wants: stable video ids, titles, channels, upload dates, language and caption source, full text, silence-grouped paragraphs (natural chunk boundaries), and per-segment timestamps for deep-linking back to the exact moment in the video.

### The cache is the scaling story
Agent workloads re-read the same sources constantly. Here, every extraction lands in a shared cache, so reads return in milliseconds without another network fetch. The first addition to an account's library costs one credit; its second, tenth, and thousandth reads are free. Daily research re-runs only pay for new library items.

### Monitoring without burning credits
The latest-uploads endpoint is free. The intended loop: poll it on a schedule, diff the video ids against what you have, extract only the new videos, hand transcripts to your summarizer. A daily channel brief costs a credit only on days the creator actually posts.

### Agents can even buy access themselves
Authentication is one string in a header: create a free API key once at /app → MCP & API → API keys, or use the license key a credit pack issues. Give your agent the key as an environment variable and it meters itself with the billing block every response includes, and tops up at the purchase_url a 402 hands it, without the key ever changing.

02:10Questions
## Questions

How do I use this from LangChain / LlamaIndex / my own code?
It is plain HTTP GET with JSON responses and an OpenAPI spec at /openapi.json. One requests.get() call, or point a codegen tool at the spec.

What about n8n, Make, or Zapier?
Use the generic HTTP request node: GET https://bulktranscripts.co/api/v1/transcript?video={{url}} with an Authorization header. The free latest-uploads endpoint makes a clean trigger.

How should I chunk transcripts for RAG?
Start from the paragraphs array. Paragraphs are grouped by natural silence gaps, which beats fixed-size chunking for spoken content. Keep the segment timestamps alongside to cite moments.

Can I run a whole channel through it?
Yes. List the channel (up to 1,000 videos), then fetch transcripts; the web app's bulk mode and the API share the same cache, so nothing is ever extracted twice.

## Start with 30 free credits
Sign in with Google, create a key, make the first call. No card.
See the integration docs
Stuck on setup? Email hello@bulktranscripts.co with the app or code you are using and what you see. A person answers, usually within a day.

## More YouTube transcript tools

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/claude-code-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in Claude Code

 One command gives Claude Code real YouTube access: summarize videos, research channels, and monitor uploads from inside your terminal.

 Setup steps ↓
 
 Works withCCodexCursorGGrok BotGChatGPTVS CodeGGemini CLICClaude CoworkPPerplexity Computer

 Claude Code · setup
claude mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp

One Google sign-in, 30 free credits.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

11 minute
### Run one command
Add the server, then type /mcp and pick Authenticate. Sign in with Google; that includes 30 free credits.
claude mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp

210 seconds
### Ask Claude Code in plain words
Claude Code picks the right tool by itself. You never name one.
Claude Code · BulkTranscripts connected
Read the last five videos on @AnthropicAI and write docs/ai-notes.md with what changed in each.
get_latest_videos15 recent uploadsfree
get_transcripts5 full transcripts5 credits
Wrote docs/ai-notes.md- One section per video, newest first
- What changed, quoted from the transcript
- A link and timestamp for every claim

00:30Setup
## Step by step

## Option 1: MCP server (recommended)
claude mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp
Run /mcp inside a session and pick Authenticate. A BulkTranscripts sign-in page opens (Google or license key; signing in includes 30 free credits). Ten tools appear once connected (get_transcript, get_transcripts, search_youtube, search_channel, get_channel_videos, get_playlist_videos, get_latest_videos, and start_bulk_extract / get_bulk_status / list_bulk_results for whole channels).
Prefer skipping OAuth? Attach a license key as a header instead:
claude mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp \
 --header "Authorization: Bearer YOUR_LICENSE_KEY"

## Option 2: agent skill
Prefer a skill Claude Code loads on demand? Install the SKILL.md once:
mkdir -p ~/.claude/skills/youtube-transcripts
curl -fsSL https://bulktranscripts.co/skill.md \
 -o ~/.claude/skills/youtube-transcripts/SKILL.md
Set BULKTRANSCRIPTS_API_KEY in your environment when you have a license key. Details: the agent skill page.

## Try asking
- Summarize this video and pull the key quotes: https://youtube.com/watch?v=…
- What has @TED published about AI? Fetch the three most relevant transcripts.
- Check if @mkbhd uploaded anything this week and brief me on it.
- Get every transcript in this course playlist and turn it into study notes.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.
If tools don't appear, start a new session (MCP servers are read at session start) and check claude mcp list.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
MCP serverclaude mcp add --transport http …Google, in the browserMost people. Ten tools, nothing to manage
MCP with a keyThe same command plus --headerAPI key or license keyHeadless boxes, CI, shared machines
Agent skillA SKILL.md in ~/.claude/skillsBULKTRANSCRIPTS_API_KEYKeeping MCP off; the skill loads only when needed

01:10Try asking
## What to ask Claude Code
Paste any of these once the server is connected.

Docs from talks“Read this conference talk and add its API changes to CHANGELOG-notes.md: https://youtube.com/watch?v=…”

Learn a library“Find the three best tutorials on LangGraph on YouTube and write a cheat sheet from them.”

Research a repo“What have the maintainers of this project said on YouTube about the roadmap?”

Course to notes“Turn every lecture in this playlist into one Markdown file per lecture in notes/.”

Weekly digest“Check @Fireship and @ThePrimeagen for new uploads and summarize anything about Rust.”

Quote check“Did Andrej Karpathy say this in his last video? Quote the exact line and timestamp.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in Claude Code or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, Claude Code runs the next block from video 1,001; anything already in your library comes back free.

01:30Which one
## Claude Code, Claude Desktop or Claude.ai?
One BulkTranscripts account works in all three. Only the setup differs.

### Claude Code
Your terminal. One claude mcp add command, or the agent skill. Best when the transcripts should end up in files or code.

### Claude.ai and Desktop
Settings → Connectors → Add custom connector with https://bulktranscripts.co/mcp. Best for reading and chatting. Connector guide

### Both at once
Connect it in each app with the same Google account. Credits and your transcript library are shared, so a video you read once is free everywhere.

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about Claude Code

How do I add BulkTranscripts to Claude Code?
Run claude mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp, then type /mcp in a session, choose Authenticate and sign in with Google. Ten YouTube tools appear once you are signed in.

Do I need an API key for Claude Code?
Not for the MCP server: the Google sign-in is the credential. You need an API key only for the agent skill or for headless setups that pass an Authorization header. Creating a key is free in the app under MCP & API.

The BulkTranscripts tools don't show up in Claude Code. What now?
Start a new session, because MCP servers are read when a session starts. Then run claude mcp list to check the server is there, and /mcp to see whether it still needs authentication.

Is BulkTranscripts free to use in Claude Code?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can Claude Code read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in Claude Code
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/codex-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in Codex

 Point Codex at one remote MCP server and it can read any YouTube video, channel, or playlist as text.

 Setup steps ↓
 
 Works withClaude CodeCursorGGrok BotGChatGPTVS CodeGGemini CLICClaude CoworkPPerplexity Computer

 Codex · setup
codex plugin marketplace add pratie/bulktranscripts-mcp
codex plugin add bulktranscripts@bulktranscripts
codex mcp login bulktranscripts

One Google sign-in, 30 free credits.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

11 minute
### Install the plugin
Add the BulkTranscripts marketplace, install the plugin, then sign in with Google once. It brings the MCP server and the skill together.
codex plugin marketplace add pratie/bulktranscripts-mcp
codex plugin add bulktranscripts@bulktranscripts
codex mcp login bulktranscripts

210 seconds
### Ask Codex in plain words
Codex picks the right tool by itself. You never name one.
Codex · BulkTranscripts connected
Read the release talk for this framework and update MIGRATION.md with the breaking changes: https://youtube.com/watch?v=…
get_transcript1 full transcript1 credit
Updated MIGRATION.md- Five breaking changes, each with the quote that announced it
- Code changes needed in this repo, file by file
- Timestamps to re-watch the tricky parts

00:30Setup
## Step by step

## Option 1: Codex plugin (recommended)
BulkTranscripts is a Codex plugin: one install brings the MCP server and a skill that tells Codex when to use each tool. Add the marketplace, install, then sign in with Google once:
codex plugin marketplace add pratie/bulktranscripts-mcp
codex plugin add bulktranscripts@bulktranscripts
codex mcp login bulktranscripts
The source is public at pratie/bulktranscripts-mcp. In the Codex app, open Settings → Plugins, choose Add more and paste https://github.com/pratie/bulktranscripts-mcp.git.

## Option 2: MCP server in config.toml
On a Codex build without plugins, add the remote (streamable HTTP) server to ~/.codex/config.toml yourself:
[mcp_servers.bulktranscripts]
url = "https://bulktranscripts.co/mcp"
Then run codex mcp login bulktranscripts to sign in with Google. On builds without MCP OAuth, use the key-in-URL form; it behaves identically:
[mcp_servers.bulktranscripts]
url = "https://bulktranscripts.co/mcp?key=YOUR_API_KEY"
Nothing to buy for that: create a free API key in the app. Sign in with Google at bulktranscripts.co/app, then MCP & API → API keys → Create key. It includes the 30 free credits and works on the MCP endpoint and the REST API alike. A license key from a credit-pack purchase works in the same slot.

## Option 3: agent skill
Codex reads Agent Skills (SKILL.md folders) too:
mkdir -p ~/.codex/skills/youtube-transcripts
curl -fsSL https://bulktranscripts.co/skill.md \
 -o ~/.codex/skills/youtube-transcripts/SKILL.md
The skill teaches Codex the plain REST API, so it works even on Codex builds without remote-MCP support. It just needs curl.

## Try asking
- Summarize this video and pull the key quotes: https://youtube.com/watch?v=…
- What has @TED published about AI? Fetch the three most relevant transcripts.
- Check if @mkbhd uploaded anything this week and brief me on it.
- Get every transcript in this course playlist and turn it into study notes.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.
On older Codex versions, verify MCP support with codex mcp --help; if your build only supports stdio servers, use the skill instead.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
Codex plugincodex plugin add bulktranscripts@bulktranscriptsGoogle, via codex mcp loginMost people. MCP server and skill in one install
MCP server[mcp_servers.bulktranscripts] in config.tomlGoogle, or a key in the URLBuilds without plugin support
Agent skillSKILL.md in ~/.codex/skillsBULKTRANSCRIPTS_API_KEYBuilds without remote MCP; only needs curl
REST APIPlain curl callsAPI keyScripts Codex writes for you

01:10Try asking
## What to ask Codex
Paste any of these once the server is connected.

Migration notes“Read this release talk and list every breaking change that affects this repo.”

Tutorial to code“Follow this tutorial video and scaffold the same project here: https://youtube.com/watch?v=…”

Library research“Find the three most-watched videos on Bun vs Node and summarize the trade-offs.”

Docs from talks“Turn this conference talk into a docs/ page with the architecture it describes.”

Changelog watch“Check @vercel for new uploads this week and note anything about Next.js.”

Quote check“Did the speaker really recommend this pattern? Quote the line and timestamp.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in Codex or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, Codex runs the next block from video 1,001; anything already in your library comes back free.

01:30Which one
## Codex or ChatGPT?
Both are OpenAI. The setup is different.

### Codex CLI
A terminal coding agent. One block in config.toml or the skill. Best when the transcript should change files.

### ChatGPT
The chat app. A custom connector in Developer mode. Best for reading, summaries and research. ChatGPT guide

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about Codex

How do I add BulkTranscripts to Codex?
Install the plugin: run codex plugin marketplace add pratie/bulktranscripts-mcp, then codex plugin add bulktranscripts@bulktranscripts, then codex mcp login bulktranscripts and sign in with Google. On builds without plugins, add [mcp_servers.bulktranscripts] with url = "https://bulktranscripts.co/mcp" to ~/.codex/config.toml instead.

My Codex build has no remote MCP support. Can I still use it?
Yes. Install the agent skill instead: it teaches Codex the plain REST API, so all it needs is curl and a free API key in BULKTRANSCRIPTS_API_KEY.

Is BulkTranscripts free to use in Codex?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can Codex read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in Codex
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/cursor-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in Cursor

 Cursor's agent can fetch transcripts, search channels, and track uploads once it knows about one URL.

 Setup steps ↓
 
 Works withClaude CodeCCodexGGrok BotGChatGPTVS CodeGGemini CLICClaude CoworkPPerplexity Computer

 Cursor · setup
{
 "mcpServers": {
 "bulktranscripts": {
 "url": "https://bulktranscripts.co/mcp"
 }
 }
}

One Google sign-in, 30 free credits.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

130 seconds
### Click Add to Cursor
The install button in the guide writes the config for you. Or paste this into ~/.cursor/mcp.json, then sign in with Google under Settings → MCP.
{
 "mcpServers": {
 "bulktranscripts": { "url": "https://bulktranscripts.co/mcp" }
 }
}

210 seconds
### Ask Cursor in plain words
Cursor picks the right tool by itself. You never name one.
Cursor · BulkTranscripts connected
Pull the transcript of this conference talk and write the architecture decisions into docs/adr/.
get_transcript1 full transcript1 credit
Created docs/adr/0007-event-sourcing.md- The decision, the context and the trade-offs
- Direct quotes from the speaker
- A link to the moment each point was made

00:30Setup
## Step by step

## One-click install
Add to Cursor

## Or configure it by hand
Global: ~/.cursor/mcp.json · per-project: .cursor/mcp.json in the repo root:
{
 "mcpServers": {
 "bulktranscripts": {
 "url": "https://bulktranscripts.co/mcp"
 }
 }
}
Under Settings → MCP the server shows a login prompt. Click it and sign in with Google or your license key. Prefer no OAuth? Append the license key to the URL instead (https://bulktranscripts.co/mcp?key=…).
Then use Agent mode; the tools are picked up by the model as needed.

## Try asking
- Research how @fireship explains WebSockets: find the video and summarize it.
- Pull the transcript of this conference talk and extract the architecture decisions.
- List everything this channel published about Rust, newest first.
- Draft a blog post from this video's transcript, keeping direct quotes.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
One-click installThe Add to Cursor buttonGoogle, in the browserFastest; nothing to edit
mcp.json~/.cursor/mcp.json or .cursor/mcp.jsonGoogle, or ?key= in the URLSharing the setup with a team repo

01:10Try asking
## What to ask Cursor
Paste any of these once the server is connected.

Talk to ADR“Pull this conference talk and write an ADR for the approach it recommends.”

Learn by video“Research how @fireship explains WebSockets and add a short explainer to README.md.”

Channel scan“List everything this channel published about Rust, newest first.”

Blog draft“Draft a blog post from this video's transcript, keeping direct quotes.”

Course notes“Turn this playlist into one Markdown note per lesson in notes/.”

Compare advice“Find two popular videos on React Server Components and list where they disagree.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in Cursor or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, Cursor runs the next block from video 1,001; anything already in your library comes back free.

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about Cursor

How do I add BulkTranscripts to Cursor?
Use the Add to Cursor button on this page, or add {"mcpServers": {"bulktranscripts": {"url": "https://bulktranscripts.co/mcp"}}} to ~/.cursor/mcp.json. Then open Settings → MCP and sign in with Google.

Which Cursor mode uses the tools?
Agent mode. The model calls the BulkTranscripts tools on its own when a request needs YouTube.

Is BulkTranscripts free to use in Cursor?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can Cursor read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in Cursor
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/grok-bot-youtube-transcripts
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Can Grok watch YouTube videos? How to make it work

 Paste a YouTube link into Grok and it usually answers from the title, the description and what others wrote about the video, not from what was said. Give it the transcript and it can summarize, quote and fact-check the real thing. Three ways to do it, from a 30-second paste to a whole channel.

 Setup steps ↓
 Published August 26, 2026 · By the BulkTranscripts team

 Works withClaude CodeCCodexCursorGChatGPTVS CodeGGemini CLICClaude CoworkPPerplexity Computer

 Grok Bot · setup
Connect the BulkTranscripts remote MCP server at https://bulktranscripts.co/mcp.
Use it whenever I ask you to research, summarize, monitor, or quote YouTube.
After connecting, show me the available tools and test the connection on this video: [PASTE A YOUTUBE URL]

One Google sign-in, 30 free credits.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

12 minutes
### Add a custom connector
In Grok open Connectors → New Connector → Custom and paste the URL. Sign in with Google on the page that opens.
https://bulktranscripts.co/mcp

210 seconds
### Ask Grok Bot in plain words
Grok Bot picks the right tool by itself. You never name one.
Grok Bot · BulkTranscripts connected
Every Monday, check what these three channels published and give me a one-page competitor brief.
get_latest_videos3 channels checkedfree
get_transcripts4 new transcripts4 credits
Monday competitor brief- What each channel shipped this week
- The two themes they share
- A source link and timestamp for every claim

00:30Setup
## Step by step

## Why a YouTube link isn't enough
Grok reads text, and a YouTube link gives it a title, a description and whatever has been written about the video elsewhere, not the audio. Ask it to summarise a link it cannot get captions for and it will often answer from that surrounding text, which reads convincingly and can be wrong. The reliable fix is to hand Grok the transcript. There are three ways, depending on which Grok you use:
- Grok chat, one video: paste the link into BulkTranscripts, copy the AI-ready text and paste it into the conversation. No setup, 30 free transcripts.
- Grok chat, many videos: add BulkTranscripts as a custom connector (steps below). Grok then fetches transcripts itself, one video or a whole playlist, and can cite the video and timestamp for each claim.
- Grok Bot or Grok Build: connect the same MCP server so a routine or a coding session can read YouTube without driving a browser.

## Browser control makes transcripts an expensive job
Grok Bot can open YouTube on its cloud computer. Reading what was said is less reliable. Caption requests can be blocked from cloud IPs, the transcript drawer can load empty, and the fallback is a slow loop of clicking, waiting, and scrolling. That burns agent usage before the analysis has even started.
A transcript service reduces that loop to one tool call. This matters even more when you want to follow five competitors, search a founder's back catalogue, or check a course against the latest documentation. BulkTranscripts gives Grok Bot clean text and timestamps without asking it to operate YouTube's interface.
The connection exposes ten tools. Grok can search YouTube, list a channel or playlist, fetch one transcript or a batch, search inside one channel, and check recent uploads. It can decide which videos matter before spending time reading them.

## Paste this into Grok Bot
Connect the BulkTranscripts remote MCP server at https://bulktranscripts.co/mcp.
Use it whenever I ask you to research, summarize, monitor, or quote YouTube.
After connecting, show me the available tools and test the connection on this video: [PASTE A YOUTUBE URL]
The Bot can handle the setup and tell you if it needs approval. Connecting opens a BulkTranscripts sign-in page. Continue with Google and the account includes 30 free transcript credits, so there is no key to paste into the conversation.

## Connect the hosted MCP server
If you prefer to connect it yourself, BulkTranscripts is a public Streamable HTTP server. There is no local process or tunnel to keep alive. The server URL is:
https://bulktranscripts.co/mcp
- Open Grok → Connectors, or visit grok.com/connectors.
- Choose New Connector → Custom.
- Name it BulkTranscripts and paste https://bulktranscripts.co/mcp as the MCP server URL.
- Finish the connection, start a fresh conversation, and ask Grok to list the available BulkTranscripts tools.
In Grok Bot, connected services live under Settings → Plugins. If the connector has no field for a header, use the key-in-URL form instead: https://bulktranscripts.co/mcp?key=YOUR_LICENSE_KEY. It behaves identically; treat the full URL as a secret.

## Grok Build setup
The terminal agent has a direct command for remote HTTP servers:
grok mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp
Run grok mcp doctor bulktranscripts if the tools do not appear. Grok Build also reads Cursor and Claude MCP configuration, which is handy if you already connected BulkTranscripts in another coding agent.

## A workflow worth saving as a routine
Start with a narrow research contract: name the channels, the time window, the question you care about, and the shape of the final answer. Ask the Bot to check titles first, fetch only likely matches, cite every claim with a video URL and timestamp, and say when captions are unavailable. Once the result looks right, save that instruction as a routine.

## Try asking
- Every Monday, check what these three channels published. Read only the relevant transcripts and give me a one-page competitor brief with source links.
- Search @ycombinator for videos about finding a first customer. Compare the advice and cite the exact moments where speakers disagree.
- Read this course playlist, group the lessons into a study plan, and keep each lesson linked to its original video.
- Find every video this creator has made about pricing. Build a timeline of how their advice changed, with quotes and timestamps.

## How this saves time
Grok first works with small video records (title, date, channel, and URL), then pulls full transcript text only for the useful candidates. That matters on a 500-video archive. You avoid feeding the model hours of irrelevant speech, and you get an answer that can still be checked against the original recording.
Transcripts already added to your library can be read again without another credit. The latest-upload check is free, so a scheduled routine only spends a credit when it finds and fetches a new transcript.

## What to know before you connect it
- Captions are the source. If a video has no accessible caption track, the tool reports it and does not charge for the failed extraction.
- Keep citations in the prompt. Summaries are easier to trust when every important claim points back to a video and timestamp.
- Do not paste paid keys into chat. Sign in with Google on the connector's authorization page for the free credits. For paid use, add the license key through the connector's authentication field or a secure MCP header.
- Review scheduled output. A persistent Bot can keep working, but it should not turn uncertain transcript text into confident claims.

## Sources and next steps
xAI documents Grok Bot's persistent computer and MCP support, custom MCP connectors in Grok, and the Grok Build MCP commands. For the transcript tool list, authentication options, and request examples, see the BulkTranscripts MCP overview or the full MCP documentation.
Explore the YouTube MCP server

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
Grok connectorConnectors → New Connector → CustomGoogle, in the browserGrok chat and Grok Bot
Grok Buildgrok mcp add --transport http …Google, or a key in the URLThe terminal coding agent
Paste the textCopy a transcript from the web appNoneOne video, no setup

01:10Try asking
## What to ask Grok Bot
Paste any of these once the server is connected.

Weekly brief“Every Monday, check these three channels and write a one-page competitor brief with source links.”

Find disagreements“Search @ycombinator for videos about finding a first customer and show where speakers disagree.”

Study plan“Read this course playlist and group the lessons into a study plan.”

Advice over time“Find every video this creator made about pricing and show how their advice changed.”

Quote it“What exactly did the CEO say about layoffs in this interview? Quote and timestamp it.”

New uploads“Has @lexfridman posted anything this week? Summarize the newest episode.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in Grok Bot or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, Grok Bot runs the next block from video 1,001; anything already in your library comes back free.

01:30Which one
## Grok, Grok Bot or Grok Build?
One BulkTranscripts account works in all of them.

### Grok chat
Add a custom connector once and ask about any video, channel or playlist.

### Grok Bot
The same connector, plus routines: save a research prompt and it runs on a schedule.

### Grok Build
The terminal agent: grok mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp.

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about Grok Bot

Can Grok watch YouTube videos?
Not from a link alone. Grok reads text, so a YouTube link only gives it the title and description. Paste the transcript in, or connect BulkTranscripts so Grok fetches it itself, and it reads what was actually said, with timestamps it can cite.

How do I connect BulkTranscripts to Grok Build?
Run grok mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp. If the tools don't appear, run grok mcp doctor bulktranscripts.

Is BulkTranscripts free to use in Grok Bot?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can Grok Bot read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in Grok Bot
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/chatgpt-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in ChatGPT

 ChatGPT can't watch videos, but with one connector it can read any video, channel, or playlist as text.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotVS CodeGGemini CLICClaude CoworkPPerplexity Computer

 ChatGPT · BulkTranscripts
https://bulktranscripts.co/mcp

Add it as a connector, sign in with Google, then just ask.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

12 minutes
### Create a custom connector
Turn on Developer mode under Settings → Connectors → Advanced, click Create, paste the URL and choose OAuth.
https://bulktranscripts.co/mcp

210 seconds
### Ask ChatGPT in plain words
ChatGPT picks the right tool by itself. You never name one.
ChatGPT · BulkTranscripts connected
Find the 15 videos on this channel that explain how they build AI agents, and turn them into one playbook.
search_channel15 relevant videos1 credit
get_transcripts15 full transcripts15 credits
Your playbook, built from all 15 transcripts- Every video and its core idea, in one table
- Rules, workflows and prompts in ten sections
- Each point traced to the video it came from

00:30Setup
## Step by step

## Add the connector
- Open Settings → Connectors → Advanced and enable Developer mode (available on paid ChatGPT plans).
- Click Create (add a custom connector).
- Name it BulkTranscripts and set the MCP server URL to https://bulktranscripts.co/mcp.
- Authentication: OAuth. Connect, and sign in with Google on the BulkTranscripts page that opens (30 free credits included).
- In a new chat, enable the connector under the tools menu and ask away.

## Using a license key instead
Credits follow your signed-in account, so top-ups apply automatically. If you prefer not to use OAuth, ChatGPT's connector form has no header field, so put the license key in the URL and pick No authentication:
https://bulktranscripts.co/mcp?key=YOUR_LICENSE_KEY

## Try asking
- Summarize this video with the five most quotable lines: [paste URL]
- Compare how these three videos explain transformers: [three URLs]
- What did @TED upload this week? Summarize anything about health.
- Get the transcripts for this lecture playlist and make me a study guide.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
Custom connectorDeveloper mode → Create, with OAuthGoogle, in the browserPaid ChatGPT plans
Connector with a keyThe same form, URL with ?key=License key in the URLSkipping OAuth
Paste the textCopy a transcript from the web appNoneFree plans, one video

01:10Try asking
## What to ask ChatGPT
Paste any of these once the server is connected.

Quotable lines“Summarize this video with the five most quotable lines: [paste URL]”

Compare explainers“Compare how these three videos explain transformers: [three URLs]”

Weekly check“What did @TED upload this week? Summarize anything about health.”

Study guide“Get the transcripts for this lecture playlist and make me a study guide.”

Podcast books“Pull every episode of this podcast and list the books the host recommends.”

Expert disagreement“Find the most-watched videos on intermittent fasting and tell me where experts disagree.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in ChatGPT or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, ChatGPT runs the next block from video 1,001; anything already in your library comes back free.

01:30Which one
## ChatGPT or Codex?
Both are OpenAI. Pick by where the result should go.

### ChatGPT
Reading, research, summaries and study guides. A custom connector in Developer mode.

### Codex
Turning a video into code or docs in your repo. One block in config.toml. Codex guide

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about ChatGPT

Does the ChatGPT connector need a paid plan?
Custom connectors need Developer mode, which ChatGPT offers on paid plans. On a free plan, paste the link into the BulkTranscripts web app and copy the transcript into ChatGPT instead.

Can ChatGPT summarize a YouTube video without a connector?
Only from the title, description and what others wrote about it, which can be wrong. With the connector it reads the real transcript and can quote it.

Is BulkTranscripts free to use in ChatGPT?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can ChatGPT read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in ChatGPT
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/vscode-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in VS Code

 Copilot's agent mode picks up MCP tools from one JSON file. Here is the exact block for YouTube transcripts.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotGChatGPTGGemini CLICClaude CoworkPPerplexity Computer

 VS Code · setup
{
 "servers": {
 "bulktranscripts": {
 "type": "http",
 "url": "https://bulktranscripts.co/mcp"
 }
 }
}

One Google sign-in, 30 free credits.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

11 minute
### Add one entry to mcp.json
Create .vscode/mcp.json or run MCP: Add Server and pick HTTP. VS Code asks you to sign in when the server starts.
{
 "servers": {
 "bulktranscripts": {
 "type": "http",
 "url": "https://bulktranscripts.co/mcp"
 }
 }
}

210 seconds
### Ask VS Code in plain words
VS Code picks the right tool by itself. You never name one.
Copilot agent mode · BulkTranscripts connected
Find the official tutorial videos for this library and summarize the setup steps into CONTRIBUTING.md.
search_youtube8 tutorial videos1 credit
get_transcripts3 full transcripts3 credits
Updated CONTRIBUTING.md- Setup steps merged from three tutorials
- Where the videos disagree, flagged
- Links to each source video

00:30Setup
## Step by step

## Workspace setup
Create .vscode/mcp.json in your project (or run MCP: Add Server from the Command Palette and pick HTTP):
{
 "servers": {
 "bulktranscripts": {
 "type": "http",
 "url": "https://bulktranscripts.co/mcp"
 }
 }
}
VS Code prompts for the server's OAuth sign-in when it first starts. Continue with Google or your license key (signing in includes the 30 free credits). Then open Copilot Chat in Agent mode, click the tools icon, and confirm the bulktranscripts tools are listed. VS Code 1.101+ supports remote HTTP servers and MCP auth natively.

## Paid credits
Use the key-in-URL form (https://bulktranscripts.co/mcp?key=YOUR_LICENSE_KEY) or add a header via the server's headers field:
{
 "servers": {
 "bulktranscripts": {
 "type": "http",
 "url": "https://bulktranscripts.co/mcp",
 "headers": {
 "Authorization": "Bearer YOUR_LICENSE_KEY"
 }
 }
 }
}

## Try asking
- Watch this talk for me and write ADR-style notes: [URL]
- Find the official tutorial videos for this library and summarize the setup steps.
- Transcribe this playlist and generate a README section from it.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
mcp.json with OAuth"type": "http" entryGoogle, in the browserMost people; VS Code 1.101+
mcp.json with a headerAdd headers.AuthorizationAPI key or license keyShared machines, Codespaces

01:10Try asking
## What to ask VS Code
Paste any of these once the server is connected.

ADR notes“Watch this talk for me and write ADR-style notes: [URL]”

Setup steps“Find the official tutorial videos for this library and summarize the setup steps.”

Playlist to README“Transcribe this playlist and generate a README section from it.”

Bug context“Did the maintainers mention this bug in a livestream? Find the moment.”

Weekly digest“Check @code for new uploads and summarize what changed in VS Code.”

Lecture notes“Turn this lecture into notes with code examples in notes/.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in VS Code or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, VS Code runs the next block from video 1,001; anything already in your library comes back free.

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about VS Code

Which VS Code version supports the BulkTranscripts MCP server?
VS Code 1.101 and later support remote HTTP MCP servers and their OAuth sign-in. Use Copilot Chat in Agent mode and check the tools icon for the bulktranscripts tools.

Is BulkTranscripts free to use in VS Code?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can VS Code read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in VS Code
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/n8n-youtube-transcripts
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# YouTube transcripts in n8n

 Three ways in: the BulkTranscripts node (verified by n8n), a plain HTTP Request node, or the MCP Client tool for n8n's AI Agent.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotGChatGPTVS CodeGGemini CLICClaude Cowork

 n8n · BulkTranscripts
https://bulktranscripts.co/mcp

Add it as a connector, sign in with Google, then just ask.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

12 minutes
### Add the verified node
In n8n Cloud, search BulkTranscripts in the nodes panel and install it (self-hosted: Settings → Community Nodes → Install), then create a BulkTranscripts credential with a free API key.
n8n-nodes-bulktranscripts

210 seconds
### Ask n8n in plain words
n8n picks the right tool by itself. You never name one.
n8n AI Agent · BulkTranscripts connected
Every morning, check my five favourite channels and post a summary of anything new to Slack.
get_latest_videos5 channels checkedfree
get_transcript2 new transcripts2 credits
Posted to #research- Two new videos, one paragraph each
- Key quotes with timestamps
- Nothing re-read: old videos cost nothing

00:30Setup
## Step by step

## BulkTranscripts node, verified by n8n (easiest)
n8n reviewed and verified the BulkTranscripts node in October 2026, so n8n Cloud users can find it straight from the nodes panel.
- n8n Cloud: on the canvas, open the nodes panel, search BulkTranscripts and install it. If it does not show up, an instance owner needs to allow verified community nodes first. Self-hosted n8n: open Settings → Community Nodes → Install and enter n8n-nodes-bulktranscripts (npm, source).
- Create a BulkTranscripts API credential and paste a free key from MCP & API → API keys.
- Add the BulkTranscripts node and pick an operation: get a transcript, list a channel or playlist, search YouTube or one channel, or get a channel's latest uploads for free. List operations output one item per video, so they connect straight into a transcript step.
- Import the ready-made new-uploads workflow: it polls a channel for free every hour and transcribes only videos it has not seen.
The node also works as a tool inside n8n's AI Agent. Prefer no install? Use the HTTP Request node below; both call the same API.

## HTTP Request node (any workflow)
- Add an HTTP Request node, method GET.
- URL: https://bulktranscripts.co/api/v1/transcript
- Query parameters: video → {{ $json.videoUrl }} (and optionally segments=0 for smaller payloads).
- Authentication → Generic → Header Auth: name Authorization, value Bearer YOUR_API_KEY.
- The response is JSON: text, paragraphs, segments, plus title/channel metadata, ready for the next node.

## The free monitoring pattern
/api/v1/channel/latest costs nothing, which makes it the perfect trigger feed:
- Schedule Trigger: e.g. every morning.
- HTTP Request: GET …/api/v1/channel/latest?channel=@yourtarget (free of credits, but send your API key like every other call).
- IF: keep items where published is newer than the last run (workflow static data works well for the cursor).
- HTTP Request: fetch the transcript for the new video ids only (1 credit each).
- Send to Slack, Notion, email, or an LLM node for a daily brief.

## MCP Client (AI Agent workflows)
Using n8n's AI Agent? Add the MCP Client Tool with endpoint https://bulktranscripts.co/mcp?key=YOUR_API_KEY and the agent gets all ten YouTube tools at once. (The MCP endpoint requires a credential and n8n doesn't run the OAuth sign-in, so use the key-in-URL form. Nothing to buy: create a free API key in the app at MCP & API → API keys and it works on both the MCP endpoint and the REST endpoints above.)

## Notes
Every REST call needs an API key. Creating one is free: sign in with Google at bulktranscripts.co/app, open MCP & API → API keys, and hit Create key. The account comes with 30 free credits, no card. Buying a one-time credit pack tops up the same account, so the key keeps working; credits never expire and repeat reads from your own library are free.
Failures (video without captions) return HTTP 4xx with a stable error.code. Branch on it with an IF node; failed extractions are never charged.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
Verified noden8n-nodes-bulktranscriptsFree API keyMost workflows; also an AI Agent tool
HTTP Request nodeGET /api/v1/transcriptHeader Auth with an API keyNo install allowed
MCP Client toolMCP Client node in an AI AgentAPI key in the URLAgents that pick tools themselves

01:10Try asking
## What to ask n8n
Paste any of these once the server is connected.

Morning digest“Schedule → latest uploads (free) → transcript → summary → Slack.”

Playlist to Notion“Playlist → one transcript per item → Notion page per video.”

Competitor watch“Five channels → new uploads only → weekly email brief.”

Search to sheet“Search YouTube for a topic → transcripts → Google Sheet of key claims.”

RAG feed“Channel videos → transcripts → chunk → vector store.”

Agent tool“AI Agent with the BulkTranscripts node as a tool, answering questions in chat.”

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about n8n

Is there an n8n node for BulkTranscripts?
Yes, and n8n has verified it. On n8n Cloud, search BulkTranscripts in the nodes panel; on self-hosted n8n, install n8n-nodes-bulktranscripts from Settings → Community Nodes. It covers transcripts, channel and playlist listings, YouTube search and free latest-upload checks, and it works as an AI Agent tool.

How do I avoid paying for the same video twice in n8n?
Poll the free latest-uploads operation and keep only new video ids (the Remove Duplicates node works well). Videos already in your library are free to read again anyway.

Is BulkTranscripts free to use in n8n?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can n8n read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in n8n
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/make-youtube-transcripts
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# YouTube transcripts in Make

 One HTTP module turns any Make scenario into a transcript pipeline, with no app approval or OAuth setup required.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotGChatGPTVS CodeGGemini CLICClaude Cowork

 Make · BulkTranscripts
https://bulktranscripts.co/mcp

Add it as a connector, sign in with Google, then just ask.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

13 minutes
### Add an HTTP module
Use HTTP → Make a request, GET the transcript endpoint and send your free API key as a Bearer header.
GET https://bulktranscripts.co/api/v1/transcript?video={{videoUrl}}
Authorization: Bearer YOUR_API_KEY

210 seconds
### Ask Make in plain words
Make picks the right tool by itself. You never name one.
Make scenario · BulkTranscripts connected
When a new video lands on my channel, transcribe it and draft a blog post in Google Docs.
channel/latestnew upload foundfree
transcript1 full transcript1 credit
Draft saved to Google Docs- Title and intro from the video
- Sections that follow the talk
- Quotes kept word for word

00:30Setup
## Step by step

## HTTP module setup
- Add HTTP → Make a request.
- URL: https://bulktranscripts.co/api/v1/transcript, method GET.
- Query string: item video = the video URL from a previous module; add segments=0 if you only need the text.
- Headers: Authorization = Bearer YOUR_API_KEY (required on every call; create a free key in the app).
- Set Parse response → Yes so title, text, and paragraphs[] map straight into later modules.

## Daily channel digest (free trigger)
- Schedule the scenario daily.
- HTTP GET …/api/v1/channel/latest?channel=@yourtarget (with the same Authorization header). This endpoint is free, so the check itself never spends credits.
- Iterator over results[] → Filter on published within the last 24h.
- HTTP GET the transcript for new videos only, then route to Google Docs, Sheets, Slack, or an AI module for summarizing.

## Try asking
- Archive every new video's transcript from three competitor channels into Sheets.
- Summarize new uploads with an AI module and post the digest to Slack.
- Append podcast transcripts to a Google Doc, one section per episode.

## Notes
Every REST call needs an API key. Creating one is free: sign in with Google at bulktranscripts.co/app, open MCP & API → API keys, and hit Create key. The account comes with 30 free credits, no card. Buying a one-time credit pack tops up the same account, so the key keeps working; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
HTTP moduleHTTP → Make a request, GETAPI key as a Bearer headerAny scenario
Free trigger feedGET /api/v1/channel/latestAPI key (no credits used)Watching channels for new uploads

01:10Try asking
## What to ask Make
Paste any of these once the server is connected.

Video to blog“New upload → transcript → AI module → Google Docs draft.”

Show notes“Podcast playlist → transcripts → show notes in Notion.”

Sheet of quotes“List of URLs in Sheets → transcripts → best quotes back into the sheet.”

Social posts“New upload → transcript → five post drafts in Airtable.”

Competitor digest“Three channels → new uploads only → weekly email.”

Course notes“Playlist → one transcript per video → study notes.”

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about Make

Does Make have a BulkTranscripts app?
Not yet. Use the HTTP module: GET https://bulktranscripts.co/api/v1/transcript with the video URL and an Authorization: Bearer header. The response is JSON that maps straight into the next module.

Is BulkTranscripts free to use in Make?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can Make read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in Make
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/gemini-cli-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in Gemini CLI

 One extension install gives Gemini CLI real YouTube access: read videos, research channels and track uploads from your terminal.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotGChatGPTVS CodeCClaude CoworkPPerplexity Computer

 Gemini CLI · setup
gemini extensions install https://github.com/pratie/bulktranscripts-mcp

One Google sign-in, 30 free credits.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

11 minute
### Install the extension
Run this in your shell. Gemini opens a Google sign-in the first time it uses a tool.
gemini extensions install https://github.com/pratie/bulktranscripts-mcp

210 seconds
### Ask Gemini CLI in plain words
Gemini CLI picks the right tool by itself. You never name one.
Gemini CLI · BulkTranscripts connected
Read the keynote from Google I/O and list every API change that affects this repo.
search_youtubethe keynote video1 credit
get_transcript1 full transcript1 credit
Seven API changes, three affect this repo- What changed, quoted from the keynote
- The files in this repo it touches
- Timestamps to re-watch each one

00:30Setup
## Step by step

## Option 1: the extension (recommended)
gemini extensions install https://github.com/pratie/bulktranscripts-mcp
Run it in your shell, not inside a Gemini session. The extension registers the BulkTranscripts server; the first time a tool runs, a sign-in page opens. Continue with Google and the account includes 30 free credits.

## Option 2: add the server yourself
gemini mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp
Or add it to settings.json under mcpServers. To use a key instead of the sign-in, add a header:
{
 "mcpServers": {
 "bulktranscripts": {
 "httpUrl": "https://bulktranscripts.co/mcp",
 "headers": { "Authorization": "Bearer YOUR_API_KEY" }
 }
 }
}

## Check it worked
Run gemini mcp list, or type /mcp in a session. If the sign-in expired, /mcp auth bulktranscripts opens it again.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
Extensiongemini extensions install …Google, in the browserMost people; one command
gemini mcp add--transport httpGoogle, in the browserChoosing your own server name
settings.jsonhttpUrl plus a headerAPI keyHeadless machines and CI

01:10Try asking
## What to ask Gemini CLI
Paste any of these once the server is connected.

Keynote notes“Read this keynote and list every announcement relevant to Android developers.”

Learn a tool“Find the best three tutorials on Firebase Genkit and write a quick-start guide.”

Repo changes“Watch this migration video and apply the same changes to this project.”

Channel digest“What did @GoogleDevelopers post this month? One line per video.”

Course notes“Turn this playlist into Markdown notes, one file per lesson.”

Fact check“Did the speaker say Gemini supports this? Quote the line and timestamp.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in Gemini CLI or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, Gemini CLI runs the next block from video 1,001; anything already in your library comes back free.

01:30Which one
## Gemini CLI or the Gemini app?
Only the CLI can load third-party MCP servers today.

### Gemini CLI
The open-source terminal agent. Supports extensions and remote MCP servers, so it can use BulkTranscripts directly.

### Gemini app
The chat app at gemini.google.com. It does not load custom MCP servers; paste a transcript from the web app instead.

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about Gemini CLI

How do I add BulkTranscripts to Gemini CLI?
Run gemini extensions install https://github.com/pratie/bulktranscripts-mcp in your shell, or gemini mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp. Gemini opens a Google sign-in the first time it uses a tool.

Can the Gemini app read YouTube transcripts through BulkTranscripts?
Not through MCP: the Gemini app does not load third-party MCP servers. Use Gemini CLI, or copy a transcript from the BulkTranscripts web app into the chat.

Is BulkTranscripts free to use in Gemini CLI?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can Gemini CLI read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in Gemini CLI
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/claude-cowork-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in Claude Cowork

 Cowork can research a channel, read a course and write the report for you once it has a way to read YouTube. That is one connector.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotGChatGPTVS CodeGGemini CLIPPerplexity Computer

 Claude Cowork · BulkTranscripts
https://bulktranscripts.co/mcp

Add it as a connector, sign in with Google, then just ask.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

12 minutes
### Add a custom connector
Customize → Connectors → Add custom connector, paste the URL, then sign in with Google.
https://bulktranscripts.co/mcp

210 seconds
### Ask Claude Cowork in plain words
Claude Cowork picks the right tool by itself. You never name one.
Claude Cowork · BulkTranscripts connected
Read the last ten videos from these three competitors and write a two-page brief in my Documents folder.
get_latest_videos3 channels checkedfree
get_transcripts10 full transcripts10 credits
Saved competitor-brief.docx- What each competitor is pushing this month
- Claims they repeat, with quotes
- Gaps you could cover, with source links

00:30Setup
## Step by step

## Add the connector
- Open Customize → Connectors and choose Add custom connector.
- Name it BulkTranscripts and paste https://bulktranscripts.co/mcp as the URL.
- Leave authentication on the automatic sign-in and save.
- Click Connect and sign in with Google on the BulkTranscripts page. That includes 30 free credits.
On a Team or Enterprise plan an Owner adds it once under Organization settings → Connectors → Add → Custom → Web; everyone else just clicks Connect.

## Using a key instead of the sign-in
Pick No sign-in and paste the URL with your key in it. Treat that URL as a secret.
https://bulktranscripts.co/mcp?key=YOUR_API_KEY

## Check it worked
In a task, open the + menu → Connectors and make sure BulkTranscripts is on, then ask for a transcript. Connector settings can't be edited after saving: to change the authentication, remove the connector and add it again.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
Custom connectorCustomize → Connectors → Add custom connectorGoogle, in the browserMost people
Organization connectorOrganization settings → ConnectorsEach member signs inTeam and Enterprise plans
Key in the URLNo sign-in, ?key= URLAPI keySkipping OAuth

01:10Try asking
## What to ask Claude Cowork
Paste any of these once the server is connected.

Competitor brief“Read the last ten videos from these three channels and write a two-page brief.”

Course to workbook“Turn this course playlist into a workbook with exercises for each lesson.”

Podcast research“Find every episode where this podcast discussed pricing and summarize the advice.”

Interview prep“Read this CEO's last five interviews and list the questions they answer badly.”

Weekly digest“Every Friday, check these channels for new uploads and save a digest.”

Quote bank“Pull the strongest quotes about remote work from these videos into a spreadsheet.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in Claude Cowork or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, Claude Cowork runs the next block from video 1,001; anything already in your library comes back free.

01:30Which one
## Cowork, Claude.ai or Claude Code?
One BulkTranscripts account works in all three.

### Claude Cowork
Long tasks that end in files: briefs, reports, spreadsheets. Add a custom connector.

### Claude.ai and Desktop
Chatting about videos. The same custom connector. Connector guide

### Claude Code
Your terminal and your repo. One command. Claude Code guide

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about Claude Cowork

How do I connect BulkTranscripts to Claude Cowork?
Open Customize → Connectors → Add custom connector, paste https://bulktranscripts.co/mcp, save, then click Connect and sign in with Google.

Does Claude Cowork need a paid plan?
Cowork itself is offered on paid Claude plans. Custom connectors are available on every plan; the free plan allows one.

Is BulkTranscripts free to use in Claude Cowork?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can Claude Cowork read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in Claude Cowork
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/perplexity-computer-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in Perplexity Computer

 Perplexity Computer is good at research jobs. With one connector it can read the videos too, not just the pages about them.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotGChatGPTVS CodeGGemini CLICClaude Cowork

 Perplexity Computer · BulkTranscripts
https://bulktranscripts.co/mcp

Add it as a connector, sign in with Google, then just ask.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

12 minutes
### Add a custom connector
Account settings → Connectors → + Custom connector, OAuth, Streamable HTTP.
https://bulktranscripts.co/mcp

210 seconds
### Ask Perplexity Computer in plain words
Perplexity Computer picks the right tool by itself. You never name one.
Perplexity Computer · BulkTranscripts connected
Research what YouTube's top fitness creators say about creatine and write a sourced report.
search_youtube12 relevant videos1 credit
get_transcripts6 full transcripts6 credits
Report: creatine, according to 6 creators- Where they agree, with quotes
- Where they disagree, and why
- Every claim linked to the video and minute

00:30Setup
## Step by step

## Add a custom connector
- Open Account settings → Connectors and click + Custom connector.
- Name: BulkTranscripts. MCP server URL: https://bulktranscripts.co/mcp.
- Authentication: OAuth. Transport: Streamable HTTP.
- Save, then sign in with Google on the BulkTranscripts page that opens (30 free credits).
Custom connectors are part of Perplexity's paid plans (Pro, Max and Enterprise).

## Check it worked
Start a Computer task and ask it to list the BulkTranscripts tools, then ask for a transcript. If the connector shows as disconnected, open it and sign in again.

## No connector on your plan?
Paste the video link into the BulkTranscripts web app, copy the AI-ready text and give it to Perplexity as a file.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
Custom connectorAccount settings → Connectors → + Custom connectorGoogle, in the browserPro, Max and Enterprise
Paste the textCopy a transcript from the web appNoneAny plan, one video

01:10Try asking
## What to ask Perplexity Computer
Paste any of these once the server is connected.

Sourced report“Research what top creators say about creatine and write a report with sources.”

Market scan“Find recent YouTube reviews of these three products and compare the complaints.”

Expert views“What have economists on YouTube said about this policy this month?”

Talk summary“Summarize this conference talk and list the papers it cites.”

Channel history“How has this creator's view on AI changed over the past year?”

Quote check“Did this politician actually say this? Find the video and timestamp.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in Perplexity Computer or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, Perplexity Computer runs the next block from video 1,001; anything already in your library comes back free.

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about Perplexity Computer

How do I add BulkTranscripts to Perplexity Computer?
Open Account settings → Connectors → + Custom connector, paste https://bulktranscripts.co/mcp, choose OAuth and Streamable HTTP, save, then sign in with Google.

Does it work in Comet?
We have not confirmed that Comet loads custom connectors. Use Perplexity Computer, or copy a transcript from the BulkTranscripts web app.

Is BulkTranscripts free to use in Perplexity Computer?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can Perplexity Computer read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in Perplexity Computer
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/manus-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in Manus

 Manus can run a whole research job on its own. Give it a clean way to read YouTube and it stops guessing from video titles.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotGChatGPTVS CodeGGemini CLICClaude Cowork

 Manus · BulkTranscripts
https://bulktranscripts.co/mcp

Add it as a connector, sign in with Google, then just ask.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

12 minutes
### Add a custom MCP server
Settings → Connectors → Custom MCP → Direct configuration, transport HTTP.
https://bulktranscripts.co/mcp

210 seconds
### Ask Manus in plain words
Manus picks the right tool by itself. You never name one.
Manus · BulkTranscripts connected
Build a spreadsheet of every product launch this channel covered in 2026, with prices mentioned.
get_channel_videos212 videos listed1 credit
start_bulk_extract40 launch videos queued40 credits
launches-2026.xlsx is ready- 40 launches, one row each
- Price as stated in the video
- A link and timestamp per row

00:30Setup
## Step by step

## Add a custom MCP server
- Open Settings → Connectors → + Add connectors and pick the Custom MCP tab.
- Choose Add custom MCP server → Direct configuration.
- Name: BulkTranscripts. Transport: HTTP. Server URL: https://bulktranscripts.co/mcp.
- Save. The first time a task uses a tool, Manus asks you to sign in; continue with Google (30 free credits).

## Using an API key instead
If the sign-in does not complete, add a custom header to the same connector:
Authorization: Bearer YOUR_API_KEY
A free key is under MCP & API → API keys in the app.

## Cue by Manus
Cue, Manus's personal-agent app, is in invite-only early access. Where it offers custom connectors, the same URL and the same sign-in apply.

## Check it worked
The connector appears under Added connectors. Start a task that mentions BulkTranscripts and watch for the tool call in the task log.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
Custom MCP, OAuthSettings → Connectors → Custom MCPGoogle, on first useMost people
Custom MCP, keySame connector plus an Authorization headerAPI keyWhen the sign-in won't complete

01:10Try asking
## What to ask Manus
Paste any of these once the server is connected.

Launch tracker“Build a spreadsheet of every product launch this channel covered this year.”

Competitor deck“Read these competitors' last five videos and make a slide deck of their pitches.”

Course workbook“Turn this course playlist into a workbook with quizzes.”

Hiring research“Watch this candidate's conference talks and summarize their expertise.”

Trend report“Search YouTube for this topic and write a trend report from the top 20 videos.”

Daily check“Every morning, check these channels for new uploads and email me a summary.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in Manus or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, Manus runs the next block from video 1,001; anything already in your library comes back free.

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about Manus

How do I connect BulkTranscripts to Manus?
Open Settings → Connectors → + Add connectors → Custom MCP → Direct configuration, set transport to HTTP and the URL to https://bulktranscripts.co/mcp, then sign in with Google when a task first uses it.

Can Manus read a whole YouTube channel?
Yes. It lists the channel, picks the relevant videos and runs a background bulk job of up to 1,000 videos, then reads the results.

Is BulkTranscripts free to use in Manus?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can Manus read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in Manus
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/openclaw-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in OpenClaw

 Your OpenClaw agent can read any YouTube video, channel or playlist, from WhatsApp, Telegram or wherever you talk to it.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotGChatGPTVS CodeGGemini CLICClaude Cowork

 OpenClaw · setup
openclaw mcp add bulktranscripts --url https://bulktranscripts.co/mcp \
 --transport streamable-http --auth oauth
openclaw mcp login bulktranscripts

One Google sign-in, 30 free credits.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

11 minute
### Add the server and sign in
Two commands. The second opens a Google sign-in.
openclaw mcp add bulktranscripts --url https://bulktranscripts.co/mcp \
 --transport streamable-http --auth oauth
openclaw mcp login bulktranscripts

210 seconds
### Ask OpenClaw in plain words
OpenClaw picks the right tool by itself. You never name one.
OpenClaw on WhatsApp · BulkTranscripts connected
Did @hubermanlab post anything new? Send me the three main takeaways.
get_latest_videos15 recent uploadsfree
get_transcript1 full transcript1 credit
New episode, three takeaways- The protocol he recommends, in one line
- The study he leans on
- Timestamps to jump to each part

00:30Setup
## Step by step

## Option 1: MCP server
openclaw mcp add bulktranscripts --url https://bulktranscripts.co/mcp \
 --transport streamable-http --auth oauth
openclaw mcp login bulktranscripts
The login opens a BulkTranscripts sign-in; continue with Google (30 free credits). Prefer a key? Set a header instead of OAuth:
openclaw mcp set bulktranscripts '{"url":"https://bulktranscripts.co/mcp","transport":"streamable-http","headers":{"Authorization":"Bearer YOUR_API_KEY"}}'

## Option 2: skill from ClawHub
openclaw skills install @pratie/bulktranscripts-youtube
Then put your key in ~/.openclaw/openclaw.json under the skill's env as BULKTRANSCRIPTS_API_KEY (run openclaw skills list for the exact entry name).
The skill teaches the agent the plain REST API, so it needs curl and a free API key in BULKTRANSCRIPTS_API_KEY. Create the key at bulktranscripts.co/app → MCP & API → API keys; it includes the 30 free credits.

## Check it worked
openclaw mcp doctor bulktranscripts --probe for the server, openclaw skills list --eligible for the skill. A static Authorization header is ignored while auth: oauth is set, and skill environment variables don't reach sandboxed containers.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
MCP serveropenclaw mcp add … --auth oauthGoogle, in the browserMost people; ten tools
MCP with a keyopenclaw mcp set with a headerAPI keyHeadless gateways
ClawHub skillopenclaw skills install @pratie/bulktranscripts-youtubeBULKTRANSCRIPTS_API_KEYKeeping MCP off

01:10Try asking
## What to ask OpenClaw
Paste any of these once the server is connected.

New uploads“Did @hubermanlab post anything new? Send me the three main takeaways.”

Morning brief“Every morning at 8, check my five channels and send one message per new video.”

Voice note to research“Find what YouTube experts say about cold plunges and give me the consensus.”

Recipe from video“Get the ingredients and steps from this cooking video.”

Course progress“I'm on lesson 4 of this playlist. Summarize lessons 5 and 6.”

Quote it“What exactly did he say about sleep at the end of this video?”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in OpenClaw or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, OpenClaw runs the next block from video 1,001; anything already in your library comes back free.

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about OpenClaw

How do I add BulkTranscripts to OpenClaw?
Run openclaw mcp add bulktranscripts --url https://bulktranscripts.co/mcp --transport streamable-http --auth oauth, then openclaw mcp login bulktranscripts and sign in with Google. Or install the skill with openclaw skills install @pratie/bulktranscripts-youtube.

Is the BulkTranscripts skill on ClawHub?
Yes, as @pratie/bulktranscripts-youtube. It uses the REST API, so it needs a free API key in BULKTRANSCRIPTS_API_KEY.

Is BulkTranscripts free to use in OpenClaw?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can OpenClaw read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in OpenClaw
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/hermes-agent-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in Hermes Agent

 Hermes runs on your machine or a server. One MCP server gives it the transcripts of any video, channel or playlist.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotGChatGPTVS CodeGGemini CLICClaude Cowork

 Hermes Agent · setup
hermes mcp add bulktranscripts --url https://bulktranscripts.co/mcp --auth oauth

One Google sign-in, 30 free credits.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

11 minute
### Run one command
Hermes opens a Google sign-in. On a server, add --flow device to the login.
hermes mcp add bulktranscripts --url https://bulktranscripts.co/mcp --auth oauth

210 seconds
### Ask Hermes Agent in plain words
Hermes Agent picks the right tool by itself. You never name one.
Hermes Agent · BulkTranscripts connected
Read this lecture series on reinforcement learning and build me a reading list from the papers it cites.
get_playlist_videos12 lectures1 credit
get_transcripts12 full transcripts12 credits
Reading list: 23 papers- Grouped by lecture
- Why the lecturer cites each one
- The minute each paper comes up

00:30Setup
## Step by step

## Add the server
hermes mcp add bulktranscripts --url https://bulktranscripts.co/mcp --auth oauth
A sign-in opens; continue with Google (30 free credits). On a headless server use the device flow: hermes mcp login bulktranscripts --flow device.

## Or in config.yaml
mcp_servers:
 bulktranscripts:
 url: "https://bulktranscripts.co/mcp"
 auth: oauth
 # or, with a key:
 # headers: {Authorization: "Bearer ${BULKTRANSCRIPTS_API_KEY}"}
The file lives at ~/.hermes/config.yaml. Run /reload-mcp in a session after editing it.

## Check it worked
hermes mcp list and hermes mcp test bulktranscripts. The tools show up as mcp_bulktranscripts_get_transcript and so on. A 401 means the token or key is wrong: sign in again.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
hermes mcp add--auth oauthGoogle, in the browserMost people
Device flowhermes mcp login … --flow deviceGoogle, on another deviceHeadless servers
config.yaml with a keyheaders: AuthorizationAPI keyScripts and containers

01:10Try asking
## What to ask Hermes Agent
Paste any of these once the server is connected.

Reading list“Read this lecture series and build a reading list from the papers it cites.”

Research memory“Remember the main claims from these five videos and tell me when a new one contradicts them.”

Paper vs talk“Compare what this researcher said in the talk with the paper's abstract.”

Daily scan“Every day, check these ML channels and summarize anything about RL.”

Notes to wiki“Turn this playlist into wiki pages, one per lecture.”

Quote“Find the exact moment this speaker defines 'agentic' and quote it.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in Hermes Agent or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, Hermes Agent runs the next block from video 1,001; anything already in your library comes back free.

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about Hermes Agent

How do I add BulkTranscripts to Hermes Agent?
Run hermes mcp add bulktranscripts --url https://bulktranscripts.co/mcp --auth oauth and sign in with Google. On a server, use hermes mcp login bulktranscripts --flow device.

Can Hermes use an API key instead of signing in?
Yes. In ~/.hermes/config.yaml, give the server a headers entry with Authorization: Bearer and your key, then run /reload-mcp.

Is BulkTranscripts free to use in Hermes Agent?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can Hermes Agent read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in Hermes Agent
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/kimi-code-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in Kimi Code

 Kimi Code reads MCP servers from one JSON file. Add BulkTranscripts and it can read any video, channel or playlist.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotGChatGPTVS CodeGGemini CLICClaude Cowork

 Kimi Code · setup
{
 "mcpServers": {
 "bulktranscripts": { "url": "https://bulktranscripts.co/mcp" }
 }
}

One Google sign-in, 30 free credits.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

11 minute
### Add one entry, then sign in
Put this in ~/.kimi-code/mcp.json, then run /mcp-config login bulktranscripts.
{
 "mcpServers": {
 "bulktranscripts": { "url": "https://bulktranscripts.co/mcp" }
 }
}

210 seconds
### Ask Kimi Code in plain words
Kimi Code picks the right tool by itself. You never name one.
Kimi Code · BulkTranscripts connected
Follow this tutorial video and build the same CLI tool in this folder.
get_transcript1 full transcript1 credit
Built cli/ from the tutorial- Same commands and flags as the video
- Two fixes where the video's code was outdated
- Timestamps for each step

00:30Setup
## Step by step

## Add the server
Edit ~/.kimi-code/mcp.json (or use /mcp-config inside Kimi Code):
{
 "mcpServers": {
 "bulktranscripts": { "url": "https://bulktranscripts.co/mcp" }
 }
}
Then sign in from a session:
/mcp-config login bulktranscripts
Prefer a key? Add "bearerTokenEnvVar": "BULKTRANSCRIPTS_API_KEY" to the server entry and export the variable.

## Check it worked
/mcp shows the connection status. A server added during a session appears in the next one. The older kimi mcp add command and ~/.kimi/mcp.json belong to the archived legacy CLI.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
mcp.json with OAuth~/.kimi-code/mcp.json + /mcp-config loginGoogle, in the browserMost people
mcp.json with a keybearerTokenEnvVarAPI keyCI and servers

01:10Try asking
## What to ask Kimi Code
Paste any of these once the server is connected.

Tutorial to code“Follow this tutorial video and build the same project here.”

Docs from talk“Read this talk and write docs/architecture.md from it.”

Library pick“Compare the top three YouTube reviews of these two libraries.”

Channel digest“What did @Fireship post this week? One line each.”

Course notes“Turn this playlist into notes, one file per lesson.”

Quote“Find where the speaker explains the caching bug and quote it.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in Kimi Code or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, Kimi Code runs the next block from video 1,001; anything already in your library comes back free.

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth
N
### NanoClaw
ncl add-mcp-server with an API key
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about Kimi Code

How do I add BulkTranscripts to Kimi Code?
Add {"mcpServers": {"bulktranscripts": {"url": "https://bulktranscripts.co/mcp"}}} to ~/.kimi-code/mcp.json, start a new session and run /mcp-config login bulktranscripts.

Does the Kimi app support BulkTranscripts?
We found no custom MCP support in the Kimi web or mobile app. Use Kimi Code, or paste a transcript from the BulkTranscripts web app.

Is BulkTranscripts free to use in Kimi Code?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can Kimi Code read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in Kimi Code
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/nanoclaw-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in NanoClaw

 NanoClaw agents live in your chats. Add one MCP server and they can read YouTube for you, from WhatsApp, Telegram, Slack or Discord.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotGChatGPTVS CodeGGemini CLICClaude Cowork

 NanoClaw · setup
ncl groups config add-mcp-server --id <group-id> \
 --name bulktranscripts --url https://bulktranscripts.co/mcp
ncl groups restart --id <group-id>

One Google sign-in, 30 free credits.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

13 minutes
### Add the server to a group
Then store a free API key in the OneCLI vault for this server and restart the group.
ncl groups config add-mcp-server --id <group-id> \
 --name bulktranscripts --url https://bulktranscripts.co/mcp

210 seconds
### Ask NanoClaw in plain words
NanoClaw picks the right tool by itself. You never name one.
NanoClaw on Telegram · BulkTranscripts connected
Summarize the newest video from @veritasium in five bullets.
get_latest_videos15 recent uploadsfree
get_transcript1 full transcript1 credit
Newest Veritasium video- The question it answers
- The experiment and the result
- The surprising part, with a timestamp

00:30Setup
## Step by step

## Add the server to a group
ncl groups config add-mcp-server --id <group-id> \
 --name bulktranscripts --url https://bulktranscripts.co/mcp
ncl groups restart --id <group-id>
NanoClaw does not run an OAuth sign-in for MCP servers, so it needs an API key. Create a free one under MCP & API → API keys and store it in the OneCLI vault as the Authorization: Bearer value for this server. Keys pasted into the URL are rejected, and --headers is meant for non-secret values only.

## Or use the skill
npx skills add pratie/youtube-transcript-skill
Run it in your NanoClaw fork and pass BULKTRANSCRIPTS_API_KEY into the container.
The skill teaches the agent the plain REST API, so it needs curl and a free API key in BULKTRANSCRIPTS_API_KEY. Create the key at bulktranscripts.co/app → MCP & API → API keys; it includes the 30 free credits.

## Check it worked
After the restart the agent has mcp__bulktranscripts__* tools. Ask it to list them.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
MCP serverncl groups config add-mcp-serverAPI key in the OneCLI vaultMost groups
Skillnpx skills add pratie/youtube-transcript-skillBULKTRANSCRIPTS_API_KEYForks that prefer skills

01:10Try asking
## What to ask NanoClaw
Paste any of these once the server is connected.

Newest video“Summarize the newest video from @veritasium in five bullets.”

Group digest“Every evening, post new uploads from our five channels to this group.”

Settle a debate“Find the video where he said this and quote it.”

Recipe“Get the ingredients from this cooking video.”

Study help“Explain lesson 3 of this playlist in simple words.”

Compare“What do these two videos disagree about?”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in NanoClaw or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, NanoClaw runs the next block from video 1,001; anything already in your library comes back free.

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
P
### Paperclip
Skill from the company Skills page

02:10Questions
## Questions about NanoClaw

How do I add BulkTranscripts to NanoClaw?
Run ncl groups config add-mcp-server --id <group-id> --name bulktranscripts --url https://bulktranscripts.co/mcp, store a free API key in the OneCLI vault as the Bearer value, then ncl groups restart --id <group-id>.

Why does NanoClaw need an API key?
NanoClaw does not run OAuth sign-ins for MCP servers, so it authenticates with a key. Keys are free and include the 30 free credits.

Is BulkTranscripts free to use in NanoClaw?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can NanoClaw read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in NanoClaw
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/integrations/paperclip-youtube
Updated: 2026-10-24

class="">

 Integration guide · 30 free credits
 
# Use YouTube transcripts in Paperclip

 Paperclip runs a company of agents. Give the research agent YouTube and it can work from what was actually said in the videos.

 Setup steps ↓
 
 Works withClaude CodeCCodexCursorGGrok BotGChatGPTVS CodeGGemini CLICClaude Cowork

 Paperclip · setup
claude mcp add --transport http --scope user bulktranscripts https://bulktranscripts.co/mcp

One Google sign-in, 30 free credits.

 30free credits to startGoogle sign-in, no card

 1,000videos per bulk joba whole channel or playlist

 $0to read a video againyour library stays free

 Nevercredits expireone-time packs, no subscription

00:20How it works
## Two steps from install to answers

12 minutes
### Add the skill
Company Skills page → paste the repo → Add, then tick it on the agent.
https://github.com/pratie/youtube-transcript-skill

210 seconds
### Ask Paperclip in plain words
Paperclip picks the right tool by itself. You never name one.
Paperclip research agent · BulkTranscripts connected
Research how our three competitors pitch their product on YouTube and file a report for the CEO agent.
search_channel3 channels searched3 credits
get_transcripts9 full transcripts9 credits
Report filed for the CEO- Each competitor's pitch in their own words
- Claims we can answer
- Source videos and timestamps

00:30Setup
## Step by step

## Option 1: add the skill (simplest)
- Open your company's Skills page.
- Paste https://github.com/pratie/youtube-transcript-skill into the source field and click Add.
- Open the agent's Skills tab and tick the YouTube transcripts skill.
- Give the agent a free API key as BULKTRANSCRIPTS_API_KEY, held as a secret.
The skill teaches the agent the plain REST API, so it needs curl and a free API key in BULKTRANSCRIPTS_API_KEY. Create the key at bulktranscripts.co/app → MCP & API → API keys; it includes the 30 free credits.

## Option 2: MCP in the runtime
Paperclip does not speak MCP itself; the agent runtime it launches does. For Claude Code, run this as the same OS user the heartbeat runs as, then finish the sign-in once with /mcp:
claude mcp add --transport http --scope user bulktranscripts https://bulktranscripts.co/mcp
Runs are headless, so the OAuth sign-in has to be completed interactively once. The openclaw_gateway adapter can't sync skills; use MCP there.

## Check it worked
claude mcp get bulktranscripts shows OAuth: configured. For the skill, the agent's Skills tab shows it ticked.

## Notes
Connecting the MCP server takes one sign-in (Google or a license key) and includes 30 free credits, no card needed. Buying a one-time credit pack gets you a license key that doubles as the API key; credits never expire and repeat reads from your own library are free.

00:50Pick a way in
## Which setup fits you
All of them share one account, one credit balance and one library.

Way inHow you connectSign-inBest for
SkillCompany Skills page → Add → tick on the agentAPI key as a secretMost companies
MCP in Claude Codeclaude mcp add --scope user …Google, once, interactivelyAgents running on Claude Code

01:10Try asking
## What to ask Paperclip
Paste any of these once the server is connected.

Competitor pitch“Research how competitors pitch on YouTube and file a report.”

Content plan“Find the top videos in our niche and propose ten video ideas we haven't covered.”

Customer voice“Read reviews of our product on YouTube and list the top complaints.”

Hiring“Watch this candidate's talks and summarize their expertise for the hiring agent.”

Weekly watch“Every Monday, check competitor channels and update the market memo.”

Training“Turn this course playlist into onboarding docs for new agents.”

01:20Whole channels
## Shorts, live streams and big channels
One background job per channel, in Paperclip or any other client.

### Shorts
“Get the transcripts of every Short on @mkbhd.” The job reads the channel's Shorts tab instead of its regular uploads.

### Past live streams
“Pull the transcripts of @NASA's live streams.” Streams that haven't started or are still running are skipped, since they have no transcript yet.

### Past 1,000 videos
A job takes up to 1,000 videos. For a bigger channel, Paperclip runs the next block from video 1,001; anything already in your library comes back free.

01:50Other agents
## Use it in your other tools too
Same account and credits everywhere. Your library follows you.

### Claude Code
One terminal command (MCP) or the agent skill
C
### Codex
One config.toml block (MCP) or the skill

### Cursor
One-click install button or mcp.json
X
### Grok Bot
MCP server with OAuth
G
### ChatGPT
Custom connector in Developer mode

### VS Code
Copilot agent mode, one mcp.json entry

### n8n
Verified node or HTTP Request

### Make
HTTP module with an API key

### Gemini CLI
One extension install, or gemini mcp add

### Claude Cowork
Custom connector, Google sign-in

### Perplexity Computer
Custom remote connector, OAuth
M
### Manus
Custom MCP connector, OAuth or Bearer key
O
### OpenClaw
openclaw mcp add, or the ClawHub skill
H
### Hermes Agent
hermes mcp add with OAuth

### Kimi Code
One mcp.json entry, then /mcp-config login
N
### NanoClaw
ncl add-mcp-server with an API key

02:10Questions
## Questions about Paperclip

How do I give a Paperclip agent YouTube transcripts?
Add https://github.com/pratie/youtube-transcript-skill on the company Skills page, tick it on the agent, and give the agent a free API key as BULKTRANSCRIPTS_API_KEY.

Does Paperclip support MCP?
Paperclip itself does not; the agent runtime it launches does. For Claude Code, add the server with claude mcp add --scope user as the heartbeat's OS user and finish the sign-in once.

Is BulkTranscripts free to use in Paperclip?
You start with 30 free credits when you sign in with Google, no card needed. One credit fetches one new transcript. Reading a transcript that is already in your library is free, and checking a channel for new uploads is free. After that, one-time credit packs that never expire; there is no subscription.

Does every YouTube video have a transcript?
No. A transcript can only be fetched when the video has captions, either uploaded by the creator or generated by YouTube. Videos without captions are skipped and never charged.

Can Paperclip read a whole channel or playlist?
Yes. Ask for a channel or playlist and the server lists the videos, then fetches their transcripts. Large jobs of up to 1,000 videos run on the server in the background and report progress, so they survive a long conversation. A bigger channel runs in blocks (videos 1 to 1,000, then 1,001 to 2,000), and the job says where the next block starts. Ask for the channel's Shorts or its past live streams and the job reads that tab instead of regular uploads.

 02:30Go deeper
## The same engine, three other ways in

### MCP server
Ten tools for any MCP client: transcripts, search, channels, bulk jobs, upload alerts.

### REST API
Plain GET requests with an OpenAPI spec, for scripts, n8n, Make and Sheets.

### Agent skill
One SKILL.md that teaches Claude Code, Codex and OpenClaw the whole API.

### Developer docs
Every endpoint, parameter, error code and credit rule in one place.

## Connect it in Paperclip
30 free credits, no card. One sign-in for the MCP connector, one free API key for the REST API.
Get your free credits

Updated October 24, 2026

---

Source: https://bulktranscripts.co/workflows/playlist-study-notes
Updated: 2026-10-24

class="doc">

Workflow · Education

# YouTube playlist → study notes

A 20-lecture course is ten hours of video and about ninety minutes of reading. This workflow turns the playlist into structured notes, one set per lecture, in the order the course intended.

Runs on the MCP server (one sign-in,
30 free credits) or the REST API · repeat reads from
your library are free.

## How it works

- Give your assistant the playlist URL.
- It lists the playlist with get_playlist_videos, order preserved, so lesson 3 stays lesson 3.
- It fetches transcripts in batches with get_transcripts (up to 20 per call).
- It writes one notes block per video: summary, key concepts, definitions, and open questions.

## Why order matters here
Course playlists build on themselves: the definition in lecture 2 is assumed by lecture 9. BulkTranscripts returns playlist videos in playlist order, so notes generated from it read as a course, not a pile. Each transcript carries the video title and link, so every note traces back to its lecture.
Agent prompt: paste into Claude, ChatGPT, or any assistant connected to the BulkTranscripts MCP server:
Get every transcript from this playlist, in order: [PLAYLIST URL]
For each video write: a 5-bullet summary, key terms with one-line definitions,
and one exam-style question. Format as Markdown with the video title and link
as the heading. Keep each video's notes under 250 words.
Or script it against the REST API:
# 1. list the playlist (1 credit)
curl "https://bulktranscripts.co/api/v1/playlist/videos?playlist=PLAYLIST_ID"
# 2. fetch each transcript (1 credit first time, repeats free)
curl "https://bulktranscripts.co/api/v1/transcript?video=VIDEO_ID&segments=0"

## What you’ll need
An MCP-capable assistant (Claude, ChatGPT, Cursor) connected to https://bulktranscripts.co/mcp (one sign-in includes 30 free credits), or a license key for the REST API. Transcripts already in your library are re-read free, so iterating on the same videos costs nothing after the first pass.

## Common questions

### How long a playlist can this handle?
get_playlist_videos lists up to 1,000 videos; the assistant then batches transcripts 20 at a time. A 30-lecture course is a few minutes of work.

### What about lectures without captions?
They are reported and skipped without failing the batch, and nothing is charged for them.

## More workflows

Competitor channels → weekly briefPlaylist → RAG corpusYouTube video → blog postNew uploads → summarized alertsChannel archive → cited quotesEpisode → show notes & chaptersWhole channel → knowledge baseOne video → a week of postsLectures → flashcards & quizzes

Updated October 24, 2026

---

Source: https://bulktranscripts.co/workflows/competitor-channel-brief
Updated: 2026-10-24

class="doc">

Workflow · Marketing

# Competitor channels → weekly brief

Watching three competitors’ channels is a part-time job. Checking what they published and reading only what matters is one scheduled prompt.

Runs on the MCP server (one sign-in,
30 free credits) or the REST API · repeat reads from
your library are free.

## How it works

- List the competitor channels you care about.
- The assistant checks each with get_latest_videos. It's always free, so the check itself costs nothing.
- It fetches transcripts only for videos matching your topics.
- It writes a brief: what each competitor said, notable claims quoted, links included.

## The trick: the check is free
get_latest_videos reads the channel’s RSS feed and never charges a credit. That makes polling economical: a daily or weekly run only spends credits when a competitor actually published something relevant. Most runs cost zero.
Agent prompt: paste into Claude, ChatGPT, or any assistant connected to the BulkTranscripts MCP server:
Check what these channels uploaded in the last 7 days:
@competitor1, @competitor2, @competitor3
For anything about [YOUR MARKET], read the transcript and summarize their
positioning in 3 bullets with one direct quote each. If nothing relevant
was published, just say so.
Or script it against the REST API:
# free: poll as often as you like
curl "https://bulktranscripts.co/api/v1/channel/latest?channel=@competitor1"
# 1 credit, only when a video is worth reading
curl "https://bulktranscripts.co/api/v1/transcript?video=NEW_VIDEO_ID&segments=0"

## What you’ll need
An MCP-capable assistant (Claude, ChatGPT, Cursor) connected to https://bulktranscripts.co/mcp (one sign-in includes 30 free credits), or a license key for the REST API. Transcripts already in your library are re-read free, so iterating on the same videos costs nothing after the first pass.

## Common questions

### Can this run on a schedule?
Yes. Use your assistant's scheduled tasks, a cron job against the REST API, or the n8n pattern in our n8n guide. The free latest-uploads check is designed for exactly this.

### How current is the upload feed?
It reads YouTube's own RSS feed, typically minutes behind a new upload, returning the latest 15 videos with publish timestamps.

## More workflows

YouTube playlist → study notesPlaylist → RAG corpusYouTube video → blog postNew uploads → summarized alertsChannel archive → cited quotesEpisode → show notes & chaptersWhole channel → knowledge baseOne video → a week of postsLectures → flashcards & quizzes

Updated October 24, 2026

---

Source: https://bulktranscripts.co/workflows/playlist-to-rag-corpus
Updated: 2026-10-24

class="doc">

Workflow · Developers

# Playlist → RAG corpus

Video knowledge bases fail because nobody re-watches videos. Chunked transcripts with timestamps make a playlist searchable, and citable back to the exact moment.

Runs on the MCP server (one sign-in,
30 free credits) or the REST API · repeat reads from
your library are free.

## How it works

- List the playlist once and store the video metadata.
- Fetch each transcript with segments included. Every segment has text, start, and duration.
- Chunk on segment boundaries (≈30–60s windows work well) and keep video_id + start as metadata.
- Embed the chunks; answers can now cite video + timestamp, linkable as watch?v=ID&t=START.

## Segments are your chunker
The transcript endpoint returns per-line segments with start times: natural semantic boundaries that beat fixed-size character windows for spoken content. Keeping video_id and start beside each chunk means every retrieval result becomes a clickable citation into the video.
Or script it against the REST API:
# list the playlist (1 credit)
curl "https://bulktranscripts.co/api/v1/playlist/videos?playlist=PLAYLIST_ID" | jq -r '.videos[].id' \
| while read id; do
 # full transcript with timestamped segments (1 credit first time)
 curl "https://bulktranscripts.co/api/v1/transcript?video=$id" > "corpus/$id.json"
 done
Agent prompt: paste into Claude, ChatGPT, or any assistant connected to the BulkTranscripts MCP server:
Fetch every transcript in this playlist with timestamps: [PLAYLIST URL]
Chunk each one into ~45-second windows on segment boundaries and output
JSONL: {text, video_id, video_title, start_seconds, url_with_timestamp}.

## What you’ll need
An MCP-capable assistant (Claude, ChatGPT, Cursor) connected to https://bulktranscripts.co/mcp (one sign-in includes 30 free credits), or a license key for the REST API. Transcripts already in your library are re-read free, so iterating on the same videos costs nothing after the first pass.

## Common questions

### Is re-running the ingestion expensive?
No. Transcripts already in your library are free on repeat reads, so re-chunking with different window sizes costs nothing after the first pass.

### What formats can I export for offline processing?
JSON with segments from the API, or SRT/VTT/CSV exports from the web app if you prefer file-based pipelines.

## More workflows

YouTube playlist → study notesCompetitor channels → weekly briefYouTube video → blog postNew uploads → summarized alertsChannel archive → cited quotesEpisode → show notes & chaptersWhole channel → knowledge baseOne video → a week of postsLectures → flashcards & quizzes

Updated October 24, 2026

---

Source: https://bulktranscripts.co/workflows/video-to-blog-post
Updated: 2026-10-24

class="doc">

Workflow · Content

# YouTube video → blog post

The difference between “write a post about this video” and a post built from the transcript is accuracy: real quotes, the speaker’s actual argument, nothing invented.

Runs on the MCP server (one sign-in,
30 free credits) or the REST API · repeat reads from
your library are free.

## How it works

- Fetch the video's transcript.
- The assistant extracts the argument structure and the 3–5 most quotable lines.
- It drafts the post in your voice, quoting the speaker verbatim where it counts.
- You edit. The facts and quotes are already right, so editing is style, not verification.

## Why transcript-first beats summary-first
Models asked to “write about” a video they have not read will confidently invent quotes. Feeding the actual transcript inverts the failure mode: the raw material is verbatim, so the draft’s claims check out and direct quotes are real, which matters the moment your post names someone.
Agent prompt: paste into Claude, ChatGPT, or any assistant connected to the BulkTranscripts MCP server:
Get the transcript of [VIDEO URL] and draft a 900-word blog post from it.
Structure: hook, the speaker's core argument in my words, 3 direct quotes
(verbatim, attributed), one counterpoint, takeaway. My audience: [WHO].
My tone: [STYLE]. End with a link to the original video.
Or script it against the REST API:
curl "https://bulktranscripts.co/api/v1/transcript?video=VIDEO_ID&segments=0"

## What you’ll need
An MCP-capable assistant (Claude, ChatGPT, Cursor) connected to https://bulktranscripts.co/mcp (one sign-in includes 30 free credits), or a license key for the REST API. Transcripts already in your library are re-read free, so iterating on the same videos costs nothing after the first pass.

## Common questions

### Does one post cost one credit?
One credit the first time that video enters your library; every redraft after that reads it free.

### Can I do this for a whole channel's archive?
Yes. List the channel with get_channel_videos, then run this workflow per video. The bulk workspace can also export everything at once.

## More workflows

YouTube playlist → study notesCompetitor channels → weekly briefPlaylist → RAG corpusNew uploads → summarized alertsChannel archive → cited quotesEpisode → show notes & chaptersWhole channel → knowledge baseOne video → a week of postsLectures → flashcards & quizzes

Updated October 24, 2026

---

Source: https://bulktranscripts.co/workflows/new-upload-alerts
Updated: 2026-10-24

class="doc">

Workflow · Automation

# New uploads → summarized alerts

“Did they post this week?” should not require opening YouTube. The check is free; credits are spent only on videos that are actually new.

Runs on the MCP server (one sign-in,
30 free credits) or the REST API · repeat reads from
your library are free.

## How it works

- A scheduled job calls the free latest-uploads endpoint for each channel.
- It compares publish timestamps against the last run's cursor.
- New videos get their transcripts fetched. That's the only credited step.
- Summaries land wherever you work: Slack, email, Notion, or a daily note.

## The cost structure is the feature
Most monitoring setups are expensive because the poll itself costs money. Here the poll is free RSS (get_latest_videos never charges), so a check every hour across ten channels costs exactly zero until something new appears. A quiet week is a free week.
Or script it against the REST API:
# cron-friendly: free check, then fetch only what's new
curl "https://bulktranscripts.co/api/v1/channel/latest?channel=@target" \
 | jq -r '.results[] | select(.published > "'$LAST_RUN'") | .id' \
 | while read id; do curl "https://bulktranscripts.co/api/v1/transcript?video=$id&segments=0"; done
Agent prompt: paste into Claude, ChatGPT, or any assistant connected to the BulkTranscripts MCP server:
Every morning, check @channel1 and @channel2 for uploads newer than 24 hours.
If there are any, fetch the transcripts and give me a 3-bullet summary per
video with a link. If nothing is new, reply with just: 'quiet day'.
Building this in n8n or Make instead? The exact node-by-node setup is in the n8n guide and Make guide.

## What you’ll need
An MCP-capable assistant (Claude, ChatGPT, Cursor) connected to https://bulktranscripts.co/mcp (one sign-in includes 30 free credits), or a license key for the REST API. Transcripts already in your library are re-read free, so iterating on the same videos costs nothing after the first pass.

## Common questions

### How many channels can I monitor?
As many as you like. The check is free per channel. Rate limits are per-minute, so space a very large list across the hour.

### Does the digest cost credits on quiet days?
No. Credits are only spent when a new video's transcript is actually fetched.

## More workflows

YouTube playlist → study notesCompetitor channels → weekly briefPlaylist → RAG corpusYouTube video → blog postChannel archive → cited quotesEpisode → show notes & chaptersWhole channel → knowledge baseOne video → a week of postsLectures → flashcards & quizzes

Updated October 24, 2026

---

Source: https://bulktranscripts.co/workflows/research-quotes-timestamps
Updated: 2026-10-24

class="doc">

Workflow · Research

# Channel archive → cited quotes

“He said it in some video” is not a citation. This workflow finds where, quotes it verbatim, and links the exact second.

Runs on the MCP server (one sign-in,
30 free credits) or the REST API · repeat reads from
your library are free.

## How it works

- Search inside the channel with search_channel for the topic.
- Fetch transcripts for the matching videos, with timestamps on.
- The assistant extracts verbatim quotes with their start times.
- Each citation renders as quote + video link + &t= timestamp, checkable by anyone.

## Verifiable beats plausible
Timestamped segments turn a claim into a checkable reference: youtube.com/watch?v=ID&t=214s drops the reader at the moment the words were said. For journalism, academic work, or just winning an argument, that is the difference between a quote and a rumor.
Agent prompt: paste into Claude, ChatGPT, or any assistant connected to the BulkTranscripts MCP server:
Search @channel for everything about [TOPIC]. Read the top 3 matching
transcripts with timestamps and give me every direct statement they made
about it. Verbatim quotes only, each with the video title and a link to
the exact timestamp. Note where statements from different videos conflict.
Or script it against the REST API:
# search inside one channel (1 credit)
curl "https://bulktranscripts.co/api/v1/channel/search?channel=@handle&q=TOPIC&limit=5"
# transcript with per-line timestamps (1 credit first time)
curl "https://bulktranscripts.co/api/v1/transcript?video=VIDEO_ID"

## What you’ll need
An MCP-capable assistant (Claude, ChatGPT, Cursor) connected to https://bulktranscripts.co/mcp (one sign-in includes 30 free credits), or a license key for the REST API. Transcripts already in your library are re-read free, so iterating on the same videos costs nothing after the first pass.

## Common questions

### How accurate are the timestamps?
They come from the video's own caption track, the same timing YouTube uses to display captions, accurate to the second.

### Can I search across many channels at once?
Use search_youtube for a keyword sweep across all of YouTube, then search_channel to go deep on specific creators.

## More workflows

YouTube playlist → study notesCompetitor channels → weekly briefPlaylist → RAG corpusYouTube video → blog postNew uploads → summarized alertsEpisode → show notes & chaptersWhole channel → knowledge baseOne video → a week of postsLectures → flashcards & quizzes

Updated October 24, 2026

---

Source: https://bulktranscripts.co/workflows/podcast-show-notes
Updated: 2026-10-24

class="doc">

Workflow · Creators

# Episode → show notes & chapters

Show notes written from memory miss half the episode. Notes written from the transcript catch every name, number, and quotable moment, with timestamps that become YouTube chapters.

Runs on the MCP server (one sign-in,
30 free credits) or the REST API · repeat reads from
your library are free.

## How it works

- Fetch the episode transcript with timestamps.
- The assistant segments the conversation into topics and marks where each begins.
- It drafts show notes: description, chapter list with timestamps, guest quotes, resources mentioned.
- Paste the chapter list into your YouTube description and the timestamps become clickable chapters.

## Chapters for free
YouTube turns a timestamp list in the description into chapter markers automatically. Since transcript segments carry real start times, the chapter list falls out of this workflow as a side effect. No more scrubbing through your own episode hunting for topic changes.
Agent prompt: paste into Claude, ChatGPT, or any assistant connected to the BulkTranscripts MCP server:
Get the transcript of my episode with timestamps: [VIDEO URL]
Produce: (1) a 2-paragraph episode description, (2) a YouTube chapter list
(MM:SS Topic) at every topic change, (3) the 5 best guest quotes with
timestamps, (4) every book/tool/person mentioned as a resources list.
Or script it against the REST API:
curl "https://bulktranscripts.co/api/v1/transcript?video=EPISODE_VIDEO_ID"

## What you’ll need
An MCP-capable assistant (Claude, ChatGPT, Cursor) connected to https://bulktranscripts.co/mcp (one sign-in includes 30 free credits), or a license key for the REST API. Transcripts already in your library are re-read free, so iterating on the same videos costs nothing after the first pass.

## Common questions

### Does this work for long episodes?
Yes. Caption tracks exist for multi-hour videos, and the segment list keeps the assistant oriented across the whole runtime.

### Can I batch my back catalog?
List your channel with get_channel_videos and run the workflow per episode; already-fetched episodes re-read free while you refine the prompt.

## More workflows

YouTube playlist → study notesCompetitor channels → weekly briefPlaylist → RAG corpusYouTube video → blog postNew uploads → summarized alertsChannel archive → cited quotesWhole channel → knowledge baseOne video → a week of postsLectures → flashcards & quizzes

Updated October 24, 2026

---

Source: https://bulktranscripts.co/workflows/channel-knowledge-base
Updated: 2026-10-24

class="doc">

Workflow · Teams

# Whole channel → knowledge base

The answer exists, in minute 34 of a video nobody will rewatch. As text, the same archive is searchable in seconds and quotable with links.

Runs on the MCP server (one sign-in,
30 free credits) or the REST API · repeat reads from
your library are free.

## How it works

- List the channel's videos (up to 1,000) with get_channel_videos.
- Batch-fetch transcripts, or run the whole channel through the bulk workspace in one job.
- Load the combined export into Notion, NotebookLM, or your internal search.
- New uploads get appended via the free latest-uploads check, so the base stays current.

## One job, not one hundred
For archives this size the bulk workspace is usually the faster path: paste the channel URL, get every available transcript as individual files or one combined, AI-ready document with per-video headers. The API path below does the same thing scriptably.
Or script it against the REST API:
# the whole archive, scripted
curl "https://bulktranscripts.co/api/v1/channel/videos?channel=@yourchannel&limit=1000" \
 | jq -r '.videos[].id' \
 | while read id; do curl "https://bulktranscripts.co/api/v1/transcript?video=$id&format=md" \
 -o "kb/$id.md"; done
Agent prompt: paste into Claude, ChatGPT, or any assistant connected to the BulkTranscripts MCP server:
List every video on @yourchannel about [PRODUCT AREA], fetch those
transcripts, and organize them into a knowledge base outline: group by
topic, note which video covers what, and flag topics covered by no video.

## What you’ll need
An MCP-capable assistant (Claude, ChatGPT, Cursor) connected to https://bulktranscripts.co/mcp (one sign-in includes 30 free credits), or a license key for the REST API. Transcripts already in your library are re-read free, so iterating on the same videos costs nothing after the first pass.

## Common questions

### What does a 200-video channel cost?
About 200 credits on first ingestion (one per new transcript). The $4.99 starter pack covers it. Keeping it current afterwards costs only what is newly published.

### Which export should I feed to AI tools?
The combined AI-ready export (one file, every transcript, per-video headers with titles and links) was designed for NotebookLM, Claude Projects, and custom GPTs.

## More workflows

YouTube playlist → study notesCompetitor channels → weekly briefPlaylist → RAG corpusYouTube video → blog postNew uploads → summarized alertsChannel archive → cited quotesEpisode → show notes & chaptersOne video → a week of postsLectures → flashcards & quizzes

Updated October 24, 2026

---

Source: https://bulktranscripts.co/workflows/video-to-social-posts
Updated: 2026-10-24

class="doc">

Workflow · Content

# One video → a week of posts

The video already contains the hooks: the sharpest lines someone said out loud. The transcript is where you find them verbatim.

Runs on the MCP server (one sign-in,
30 free credits) or the REST API · repeat reads from
your library are free.

## How it works

- Fetch the transcript of the video.
- The assistant pulls the strongest verbatim lines, the ones that stop a scroll.
- It drafts an X thread and a LinkedIn post, each built around a real quote.
- Every post links back to the video, timestamped to the quoted moment.

## Verbatim is the edge
Generic repurposing tools summarize; summaries are forgettable. The lines that work on social are the ones actually said: specific, voiced, quotable. Working from the transcript means every hook is real and every attribution survives a community-notes check.
Agent prompt: paste into Claude, ChatGPT, or any assistant connected to the BulkTranscripts MCP server:
Get the transcript of [VIDEO URL]. Find the 5 most quotable verbatim lines.
Write: (1) an X thread with a hook from the best quote, 6 posts, each under 260
chars, ending with the timestamped video link; (2) a LinkedIn post with one
quote as the opener, 150 words, my take added: [YOUR ANGLE].
Or script it against the REST API:
curl "https://bulktranscripts.co/api/v1/transcript?video=VIDEO_ID"

## What you’ll need
An MCP-capable assistant (Claude, ChatGPT, Cursor) connected to https://bulktranscripts.co/mcp (one sign-in includes 30 free credits), or a license key for the REST API. Transcripts already in your library are re-read free, so iterating on the same videos costs nothing after the first pass.

## Common questions

### Can I run this across a whole channel weekly?
Pair it with the new-upload alerts workflow: the free check finds what's new, and only fresh videos spend a credit before drafting.

### Do repeat drafts cost extra?
No. Once a video is in your library, every re-read while you iterate on angles is free.

## More workflows

YouTube playlist → study notesCompetitor channels → weekly briefPlaylist → RAG corpusYouTube video → blog postNew uploads → summarized alertsChannel archive → cited quotesEpisode → show notes & chaptersWhole channel → knowledge baseLectures → flashcards & quizzes

Updated October 24, 2026

---

Source: https://bulktranscripts.co/workflows/lecture-flashcards-quiz
Updated: 2026-10-24

class="doc">

Workflow · Education

# Lectures → flashcards & quizzes

Rewatching lectures is the least effective revision there is. Testing yourself on what was actually taught, generated from the transcript, is the most.

Runs on the MCP server (one sign-in,
30 free credits) or the REST API · repeat reads from
your library are free.

## How it works

- Fetch transcripts for the lecture (or the whole playlist, in order).
- The assistant extracts testable facts, definitions, and derivations.
- It writes Q&A pairs: cloze deletions for definitions, open questions for concepts.
- Export as CSV for Anki, or keep it in chat as an interactive quiz with timestamped answers.

## Grounded questions only
Flashcards invented by a model from its general knowledge test the model’s syllabus, not your course’s. Cards generated from the lecture transcript test exactly what your lecturer said, and each answer can cite the minute it was explained, so a failed card links straight to the remedial two minutes of video.
Agent prompt: paste into Claude, ChatGPT, or any assistant connected to the BulkTranscripts MCP server:
Get the transcripts for this lecture playlist in order: [PLAYLIST URL]
For each lecture, write 8 flashcards: 4 cloze-deletion cards for key
definitions, 4 open questions on the concepts. Include the answer and the
timestamp where it was taught. Output as CSV: front,back,source_link.
Or script it against the REST API:
curl "https://bulktranscripts.co/api/v1/playlist/videos?playlist=COURSE_PLAYLIST_ID"
curl "https://bulktranscripts.co/api/v1/transcript?video=LECTURE_VIDEO_ID"

## What you’ll need
An MCP-capable assistant (Claude, ChatGPT, Cursor) connected to https://bulktranscripts.co/mcp (one sign-in includes 30 free credits), or a license key for the REST API. Transcripts already in your library are re-read free, so iterating on the same videos costs nothing after the first pass.

## Common questions

### Can I import the output into Anki?
Yes. Ask for CSV with front,back columns and import it directly; the timestamped source link fits on the back of the card.

### Half my course has no captions. What happens?
Videos without any caption track are reported and skipped without charge; cards are generated from the lectures that have them.

## More workflows

YouTube playlist → study notesCompetitor channels → weekly briefPlaylist → RAG corpusYouTube video → blog postNew uploads → summarized alertsChannel archive → cited quotesEpisode → show notes & chaptersWhole channel → knowledge baseOne video → a week of posts

Updated October 24, 2026

---

Source: https://bulktranscripts.co/about
Updated: 2026-09-04

class="doc">

About

# About BulkTranscripts

BulkTranscripts is an independent, one-person product that downloads every available
transcript from a YouTube channel, playlist, or video in one job, and exposes the same engine to
AI agents as a REST API, an MCP server, and an agent skill.

## Who builds it

BulkTranscripts is built and run by @attenisalluneed,
an independent developer, since July 2026. There is no company behind it, no investors, and no
sales team: the person who answers support email is the person who wrote the code. The
agent skill is open on GitHub.

## Why it exists

Most transcript tools handle one video at a time. That is fine for a single clip and useless
for an archive: a podcast back-catalogue, a lecture series, a competitor's channel, a playlist
you want to feed into NotebookLM or a RAG pipeline. BulkTranscripts was built to do the whole channel
in one job, with clean text (no caption clutter), timestamps kept where they help, and exports
that AI tools can read directly.

## How it is run

- No subscription. Credits are one-time packs and never expire. The first 30
transcripts are free with no card.

- No lock-in. Every transcript you extract can be exported in TXT, Markdown,
SRT, VTT, CSV, JSON or a combined AI-ready file, as often as you like, free.

- Honest failures. Videos without captions are reported and never charged.

- Documented for machines. The API docs,
OpenAPI spec, llms.txt and
agent skill are all public.

## Where else to find it

X / Twitter ·
GitHub ·
MCP server ·
Integration guides ·
Contact

BulkTranscripts is not affiliated with, endorsed by, or sponsored by YouTube or Google.

Updated September 4, 2026