Integrations

TranscriptFetch exposes a Model Context Protocol server over Streamable HTTP at https://transcriptfetch.com/mcp. Point Claude, Cursor, ChatGPT, VS Code, or any MCP client at it and your assistant can fetch transcripts and discover videos mid-conversation. It authenticates with the same API key as the REST API and charges the same 1 credit per successful fetch.

The server advertises four tools:

Claude (web) — one-click connector

1. In Claude.ai, open Settings → Connectors and click “Add custom connector”.
2. Set the name to TranscriptFetch and the server URL to https://transcriptfetch.com/mcp.
3. Click “Connect”. You’re taken to TranscriptFetch’s sign-in / authorization page — sign in (or create an account) and approve access.
4. Acknowledge that Claude will use your TranscriptFetch credits when it fetches (1 credit per successful fetch).
5. Done. The four tools (get_transcript, search_videos, list_channel_videos, list_playlist_videos) are now available — just paste a YouTube URL or ask Claude to find or summarize videos.

Claude Code / Desktop — API key

For Claude Code, add the server with your API key as a bearer header:

claude mcp add --transport http transcriptfetch https://transcriptfetch.com/mcp \
  --header "Authorization: Bearer tf_live_YOUR_KEY"

For Claude Desktop, bridge the HTTP server with mcp-remote in your claude_desktop_config.json:

{
  "mcpServers": {
    "transcriptfetch": {
      "command": "npx",
      "args": [
        "mcp-remote",
        "https://transcriptfetch.com/mcp",
        "--header",
        "Authorization: Bearer tf_live_YOUR_KEY"
      ]
    }
  }
}

Cursor / VS Code — API key

Add an entry to your MCP config (Cursor: ~/.cursor/mcp.json, VS Code: .vscode/mcp.json):

{
  "mcpServers": {
    "transcriptfetch": {
      "url": "https://transcriptfetch.com/mcp",
      "headers": {
        "Authorization": "Bearer tf_live_YOUR_KEY"
      }
    }
  }
}

The current stable API is served under /api/v1. We add fields without bumping the version; any breaking change ships as a new version path (/api/v2) so your integration keeps working. The original unversioned endpoints (/api/transcripts and /api/transcripts/batch) remain available as a legacy alias with their original response shape. New integrations should target /api/v1.

# base: https://transcriptfetch.com
POST /api/v1/transcripts/video      # one video's transcript
POST /api/v1/transcripts/channel    # a channel's videos
POST /api/v1/transcripts/playlist   # a playlist's videos
POST /api/v1/transcripts/search     # keyword search
POST /api/v1/transcripts/batch      # up to 50 transcripts at once

A machine-readable OpenAPI 3.1 description is available for SDK generators, Postman, and Swagger UI:

GET https://transcriptfetch.com/api/v1/openapi.json

Changelog