Flexible output
Request readable TXT with optional timestamps, or standards-compatible SRT and VTT subtitle content.
Connect Claude and other MCP clients to TranscriptWave to retrieve public YouTube transcripts in TXT, SRT, or VTT format.
https://transcriptwave.com/api/mcpThe server exposes get_youtube_transcript. Give it a public YouTube URL or video ID and it returns video metadata, available caption languages, and the selected transcript.
Request readable TXT with optional timestamps, or standards-compatible SRT and VTT subtitle content.
Connect directly over Streamable HTTP. There is no local package, Docker container, or legacy SSE transport to configure.
Only supported YouTube links are accepted. Tool calls use the same caching and per-client rate limiting as the web app.
In Claude, open Settings, choose Connectors, then select Add custom connector.
Name it TranscriptWave and enter https://transcriptwave.com/api/mcp. No OAuth credentials or custom headers are required.
Ask Claude to retrieve or summarize a public YouTube transcript. Claude requests tool approval according to your client permissions.
claude mcp add --transport http --scope user \
transcriptwave https://transcriptwave.com/api/mcp
claude mcp list{
"mcpServers": {
"transcriptwave": {
"type": "http",
"url": "https://transcriptwave.com/api/mcp"
}
}
}Choose HTTP or Streamable HTTP in your client, not the legacy SSE transport.
get_youtube_transcript| Argument | Required | Description |
|---|---|---|
url | Yes | A YouTube URL or 11-character video ID. |
language | No | A caption language code such as en or pt-BR. The first available language is used by default. |
format | No | txt (default), srt, or vtt. |
include_timestamps | No | Include [M:SS] timestamps in TXT output. Defaults to true. |
Get the English transcript for https://youtu.be/dQw4w9WgXcQ and summarize its main points.
Paste a public YouTube URL to search, copy, and download its available captions directly.
Get a transcript