BulkTranscripts MCP¶
One URL gives an MCP client real YouTube access: transcripts, search, channel and playlist archives, and new-upload tracking, with no install and no signup. BulkTranscripts is a hosted streamable-HTTP MCP server at bulktranscripts.co/mcp that rides the same production extraction chain as the BulkTranscripts web app (residential proxy routing, proof-of-origin tokens, retries, and a shared cache serving popular videos instantly). Anonymous connections get 50 free transcript extractions per IP; cached transcripts and re-exports never cost a credit, and the new-uploads tool is free forever.
Server type: Remote (Streamable HTTP, hosted)
Auth: Keyless to start; optional license key as Bearer header (or ?key= parameter)
Endpoint: https://bulktranscripts.co/mcp
Tools: 7 (live-probed; server v1.0.0, protocol 2025-03-26)
Pricing: 50 free transcript extractions per IP, then one-time credit packs (no subscription)
Built by: BulkTranscripts (bulktranscripts.co)
Why This Matters for Operators¶
Most YouTube MCP servers on GitHub run yt-dlp locally and die on YouTube bot checks within a week of setup. BulkTranscripts is maintained infrastructure: a hosted server with proxy rotation, retries and a shared cache, so transcripts keep flowing while you do other work. For operators, that converts video content from a black box into queryable business data: competitor channel monitoring, creator research, market intelligence from talks and interviews, and batch notes from courses and playlists.
The workflow the toolset enables is deliberately agentic: list a channel archive (up to 1,000 videos), pick candidates, then read only the transcripts that matter. A one-line MCP add in Claude Code, Cursor, VS Code or ChatGPT developer mode is the entire install.
Tools & Capabilities¶
| Tool | Purpose |
|---|---|
| get_transcript | Full transcript of one YouTube (or TikTok) video as clean text with metadata: title, channel, duration, upload date |
| get_transcripts | Batch fetch up to 20 videos per call from YouTube URLs or ids (1 credit per uncached video) |
| search_youtube | Keyword search across videos, channels or playlists (type parameter) |
| search_channel | Topic search inside one channel's uploads, for researching what a creator has said |
| get_channel_videos | List a channel's videos (id, title, duration, URL) without fetching transcripts; accepts @handle, channel URL or id |
| get_playlist_videos | List every video in a playlist, in playlist order (courses and series keep their sequence) |
| get_latest_videos | Newest uploads (up to 15) with publish dates from YouTube RSS; always free, no credit charged |
Seven tools confirmed by live probe (the marketing page lists six; the probe is the ground truth). The batch tool costs 1 credit per uncached video; already-cached videos return instantly and free.
Installation¶
{
"mcpServers": {
"bulktranscripts": {
"type": "http",
"url": "https://bulktranscripts.co/mcp"
}
}
}
For the Claude Code CLI: claude mcp add --transport http bulktranscripts https://bulktranscripts.co/mcp
Configuration¶
Start keyless: anonymous connections get 50 free transcript extractions per public IP with no account and no credit card. When the free allowance runs out, buy a one-time credit pack and pass its license key as a Bearer header (Authorization: Bearer YOUR_KEY), or as a ?key= parameter for clients that cannot send headers. There is no subscription tier.
Example Prompts¶
- "Summarize this talk with key quotes."
- "Turn this course playlist into study notes."
- "Watch these three competitor channels and brief me on new uploads each morning."
- "List everything this creator has said about our category across their channel archive."
Business Relevance¶
Video is where competitor positioning, creator partnerships and market shifts show up first, and transcript data is the cheapest way to index it. BulkTranscripts turns that into a zero-infrastructure subscription to YouTube's text layer: competitor channel monitoring, conference talk distillation, playbook extraction from courses, and creator research before outreach. The free-latest-uploads tool doubles as a lightweight "what did this channel post this week" monitor that never costs a credit.
Integration with CorpusIQ¶
CorpusIQ answers questions about business data: what campaigns earned, which channels convert, where revenue comes from. BulkTranscripts supplies the external market and content layer those answers live next to: an agent can pull a competitor's newest videos (BulkTranscripts), extract positioning and tool mentions from the transcripts, and cross-check claims against live business metrics in CorpusIQ. BulkTranscripts covers YouTube content; CorpusIQ's 40+ connectors remain the read-side authority for the business systems themselves.
Limitations¶
- Credit metering is IP-based for the free tier, so multiple agents behind one NAT share the 50-extraction allowance.
- TikTok transcript support exists on get_transcript but the archive and search tools are YouTube-only.
- No write path: the server reads YouTube data only, no comments API, no posting surface.
- Requires the license key for sustained volume; there is no subscription, so heavy use means repeated credit packs.
See Also¶
- External MCP Server Catalog - curated third-party MCP servers for operators
- tube-bridge MCP - self-hosted YouTube research with local semantic search
- viral-outliers MCP - viral social outlier database with transcripts
- MCP Integration Guide - connecting MCP servers to Hermes Agent