# SpeakNotes MCP Server

> Use the SpeakNotes MCP server to let AI agents transcribe and summarize media through a standard MCP interface.

## Quick setup

1. Open MCP setup page: `https://speaknotes.io/settings/mcp`
2. Create a SpeakNotes API key in `/settings/api-keys` (Pro required)
3. Add the MCP endpoint and API key header to your MCP client config
4. Call `list_speaknotes_styles` to validate the connection

## Endpoint details

- URL: `https://api.speaknotes.io/mcp`
- Transport: Streamable HTTP
- Protocol: Model Context Protocol (MCP)
- Auth: `Authorization: Bearer <API_KEY>`

## Available Tools

- `list_speaknotes_styles`: List supported summary styles
- `transcribe_audio_url`: Transcribe audio from a public URL
- `summarize_text`: Summarize raw transcript text with a selected style
- `summarize_youtube`: Summarize a YouTube video URL
- `transcribe_and_summarize_audio_url`: End-to-end transcript + summary in one call

## Typical Agent Flow

1. Send media URL to `transcribe_audio_url`
2. Pass transcript into `summarize_text` with style `meeting-notes` or `bulletpoints`
3. Return title, summary, and structured content to the user

## Related Pages

- [MCP Integration Landing Page](https://speaknotes.io/integrations/mcp)
- [MCP Setup Page](https://speaknotes.io/settings/mcp)
- [OpenClaw Integration Landing Page](https://speaknotes.io/integrations/openclaw)
- [SpeakNotes Contact](https://speaknotes.io/contact)

---
*Request this page with `Accept: text/markdown` header for clean Markdown.*