# Audio Transcription MCP

> Audio Transcription MCP is an open-source MCP server for design and media. Local MCP server that transcribes audio and video files with Whisper or OpenAI. It runs locally on your machine and is started by your AI client.

- **Kind:** MCP server
- **Pricing:** Open source
- **Platforms:** CLI, self-hosted, macOS, Windows, Linux
- **Website:** https://mp3totext.io
- **Repository:** https://github.com/cykk/audio-transcription-mcp
- **MCP Registry name:** io.github.cykk/audio-transcription-mcp
- **Categories:** design and media, files and documents, AI models and LLM tooling
- **Listing updated:** 2026-11-14

## About

This Python package exposes one tool, transcribe_file, which takes the path of a local audio or video file and returns JSON with the detected language, the full text and timestamped segments. You can pass optional model, language, task and prompt arguments.

It runs over stdio and needs Python 3.11 or newer plus ffmpeg on your PATH. Without an API key it transcribes locally with Whisper, which may download the chosen model on first use and needs enough disk space and memory. If you set OPENAI_API_KEY it sends the audio to the OpenAI transcription API instead.

The README warns that the server reads local files, so it should only be configured in clients you trust. It is a separate project from the MP3ToText web service and does not use it.

## Features

- **Local Whisper mode:** Transcribes on your machine when no API key is set.
- **OpenAI API mode:** Uses the OpenAI transcription API when OPENAI_API_KEY is provided.
- **Timestamped segments:** Returns detected language, full text and per-segment timestamps.
- **Audio and video input:** Accepts local audio and video files through ffmpeg.
- **Task and prompt options:** Optional arguments for model, language, task and prompt.

## Good for

- Transcribing a recorded meeting from inside an MCP client
- Getting timestamped text from an interview file
- Running private transcription locally without sending audio out
- Checking what language a recording contains

## Install

### Claude Code (Terminal)

```sh
claude mcp add audio-transcription-mcp -- uvx audio-transcription-mcp
```

### Claude Desktop (claude_desktop_config.json)

```json
{
  "mcpServers": {
    "audio-transcription-mcp": {
      "command": "uvx",
      "args": [
        "audio-transcription-mcp"
      ]
    }
  }
}
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "audio-transcription-mcp": {
      "command": "uvx",
      "args": [
        "audio-transcription-mcp"
      ]
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "audio-transcription-mcp": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "audio-transcription-mcp"
      ]
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "audio-transcription-mcp": {
      "command": "uvx",
      "args": [
        "audio-transcription-mcp"
      ]
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "audio-transcription-mcp": {
      "command": "uvx",
      "args": [
        "audio-transcription-mcp"
      ]
    }
  }
}
```

## Tools

- `transcribe_file`: Transcribes a local audio or video file and returns language, text and timestamped segments.

## FAQ

### What does it need installed?

Python 3.11 or newer and ffmpeg on your PATH, plus the local or api extra.

### Is audio sent to OpenAI?

Only if you provide an API key; otherwise Whisper runs locally.

### What licence is it?

MIT.

### Does it use the MP3ToText service?

No. The README says it runs independently of that product.

### Is Audio Transcription MCP free?

Yes. Audio Transcription MCP is open source, so you can use and self-host it for free.

### How do I add Audio Transcription MCP to Claude Code?

Run this command in your terminal: claude mcp add audio-transcription-mcp -- uvx audio-transcription-mcp

### What are the best alternatives to Audio Transcription MCP?

The closest alternatives to Audio Transcription MCP are HyperFrames by HeyGen, ffmpeg-skill, Claude Code Ultimate Guide, ASTra and AvatarSay — Fictional Character & Talking Video MCP.

## Alternatives

- [HyperFrames by HeyGen](https://ai-powered.com/mcp/hyperframes-by-heygen): HyperFrames by HeyGen is a freemium MCP server for design and media.
- [ffmpeg-skill](https://ai-powered.com/mcp/ffmpeg-skill): ffmpeg-skill is an open-source MCP server for design and media.
- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [ASTra](https://ai-powered.com/mcp/astra): ASTra is a freemium MCP server for files and documents.
- [AvatarSay — Fictional Character & Talking Video MCP](https://ai-powered.com/mcp/avatarsay-fictional-character-and-talking-video-mcp): AvatarSay — Fictional Character & Talking Video MCP is a freemium MCP server for design and media.
- [ChartAnt](https://ai-powered.com/mcp/chartant): ChartAnt is a freemium MCP server for design and media.
- [VfxSmith](https://ai-powered.com/mcp/vfxsmith): VfxSmith is a freemium MCP server for design and media.
- [CalderFlow](https://ai-powered.com/mcp/calderflow): CalderFlow is a free MCP server for design and media.

---
Source: https://ai-powered.com/mcp/audio-transcription-mcp
