A transcription tool your agent can call

Connect Claude.ai, ChatGPT, Claude Code, Cursor, VS Code or any MCP client. Your agent can then transcribe audio and video, read the transcript, and hand back subtitle files and a finished PDF.

Your first transcription is free · then 5 minutes a month · no card

Claude.ai and ChatGPT (sign in, no key)

It is a remote MCP server, so there is nothing to install. You need an AudioTranscribr account first (free, no card); the sign-in page cannot create one.

In Claude.ai or Claude Desktop open Settings → Connectors → Add custom connector. In ChatGPT, add a connector in developer mode.
Enter the URL https://api.audiotranscribr.com/mcp.
Sign in with your account email. We send a one-time code; there is no password.
Ask for a transcript. The same plan, minutes and limits apply as everywhere else.

Claude Code, Cursor and VS Code (API key)

These take the URL https://api.audiotranscribr.com/v1/mcp and your API key in a header. Create keys in the dashboard (several, named, revocable), or get your first one on the pricing page. The -s user flag makes it available in every folder; without it, Claude Code registers the server for the current folder only. Claude Code can also sign in instead of using a key: add https://api.audiotranscribr.com/mcp with the same command (no --header) and run /mcp to sign in.

Claude Code

claude mcp add --transport http -s user audiotranscribr \
  https://api.audiotranscribr.com/v1/mcp --header "x-api-key: YOUR_KEY"

Cursor — ~/.cursor/mcp.json

{
  "mcpServers": {
    "audiotranscribr": {
      "url": "https://api.audiotranscribr.com/v1/mcp",
      "headers": { "x-api-key": "YOUR_KEY" }
    }
  }
}

VS Code — .vscode/mcp.json

{
  "servers": {
    "audiotranscribr": {
      "type": "http",
      "url": "https://api.audiotranscribr.com/v1/mcp",
      "headers": { "x-api-key": "YOUR_KEY" }
    }
  }
}

What the agent gets

transcribe_url — transcribe media at a public URL. Waits for the result and returns the transcript text plus download links for JSON, SRT, VTT, TXT and PDF, in one call.
create_upload — get an upload link for a file on the local machine.
get_transcription — status, text and fresh download links for a transcription.
get_transcript_text — read a long transcript (or its SRT / VTT subtitles) page by page.
list_transcriptions, get_account, delete_transcription — history, minutes remaining, cleanup.

Then just ask

Example prompts

Transcribe https://audiotranscribr.com/samples/clip-22-451-01.mp3 and list the action items by speaker.

Make subtitles for ./demo.mp4 and save them next to it as demo.srt.

Transcribe this interview and give me the PDF link to send to the client.

Long recordings keep running after the first call returns; the agent is told to check back with the job id, so nothing times out.

Pricing

Your first transcription is free, then 5 minutes a month. Paid plans start at $9 a month for 10 hours of audio. All plans →