MCP Server
Connect FetchTranscript to Claude, ChatGPT, Cursor and Claude Code via the Model Context Protocol.
Connect from Claude.ai or ChatGPT (no install)
Add FetchTranscript as a custom connector with this URL, then sign in with your FetchTranscript account and allow access. It works on the web, desktop and mobile apps of the same account, with no API key to copy and nothing to install.
https://api.fetchtranscript.com/mcpCalls use your credits exactly like the API: 1 credit per successful call (comments: 1 per 100), failed calls are free. See or disconnect your connected apps in Dashboard → Connected apps.
Claude (claude.ai, Desktop, mobile)
- Open Customize → Connectors.
- Click + Add, then Add custom connector.
- Name it FetchTranscript, paste the URL above as the remote MCP server URL, click Continue.
- Keep the authentication settings Claude detects (OAuth sign-in) and click Add.
- Sign in with your FetchTranscript account and click Allow.
- In a chat, open + → Connectors and turn FetchTranscript on.
The Free plan allows one custom connector. On Team and Enterprise an owner adds it in Organization settings → Connectors, then each member connects their own account.
ChatGPT
- Open chatgpt.com/plugins on the web.
- Click the plus button, then Add custom MCP server.
- Name it FetchTranscript and paste the URL above as the Server URL under Connection.
- Choose OAuth as authentication, confirm I understand and want to continue, then Create as a plugin.
- Install it, sign in with your FetchTranscript account and click Allow.
- In a new chat, type @ and pick FetchTranscript.
Your account or workspace must allow custom MCP servers; workspace admins can restrict them.
Any MCP client that supports remote servers with OAuth (Streamable HTTP) can use the same URL, for example claude mcp add --transport http fetchtranscript https://api.fetchtranscript.com/mcp in Claude Code.
What is MCP?
The Model Context Protocol (MCP) lets AI assistants call external tools. With the FetchTranscript MCP server, your AI can fetch YouTube transcripts, search videos, and access channel data directly.
Local server with an API key (for developers)
Prefer to run the server on your machine? The open-source package fetchtranscript-mcp runs with uvx and uses an API key from your dashboard. Same tools as the connector:
// claude_desktop_config.json
{
"mcpServers": {
"fetchtranscript": {
"command": "uvx",
"args": ["fetchtranscript-mcp"],
"env": {
"FETCHTRANSCRIPT_API_KEY": "yt_your_api_key"
}
}
}
}Available Tools
Once connected, your AI assistant can use these 8 tools:
get_transcriptFetch video transcript (text, JSON, or SRT)list_languagesList available transcript languagesget_video_metadataGet video title, views, likes, durationget_chaptersGet video chapter list with timestampsget_commentsFetch top comments for a videoget_channel_infoGet channel metadata and statslist_channel_videosList videos from a channelsearch_videosSearch YouTube by queryExample Prompts
Once the MCP server is connected, try asking your AI:
“Get the transcript of YouTube video dQw4w9WgXcQ”
“Summarize this YouTube video: https://youtube.com/watch?v=...”
“Search YouTube for "machine learning explained" and summarize the top result”
“List all videos from @mkbhd and get the transcript of the latest one”
“What languages are available for this video?”