Ffvoice

Overview

Ffvoice is an MCP server that performs offline speech-to-text transcription and speaker diarization entirely on your device, with no cloud dependencies. It lets AI assistants process audio files locally and invoke transcription as a native tool during conversation.

The server runs on your machine via the uvx package manager with the command uvx ffvoice. It communicates with MCP clients over stdio and requires no authentication or API keys to get started. Version v0.8.3 is available now under the MIT license, with source code at chicogong/ffvoice-engine on GitHub.

Ffvoice sits in the AI & Agents category and integrates with MCP clients including Claude Desktop, Claude Code, Cursor, VS Code, Windsurf, and Zed. If you maintain this project, you can claim this listing to verify ownership and keep the information accurate.

Installation

// claude_desktop_config.json
{
  "mcpServers": {
    "ffvoice": {
      "command": "uvx",
      "args": [
        "ffvoice"
      ]
    }
  }
}