{
  "markdown": "<p align=\"center\">\n  <img src=\"assets/logo.png?v=2\" alt=\"Waxberry\" width=\"80\" />\n</p>\n\n# live-translate-mcp\n\n[![live-translate-mcp MCP server](https://glama.ai/mcp/servers/waxberry-dev/live-translate-mcp/badges/score.svg)](https://glama.ai/mcp/servers/waxberry-dev/live-translate-mcp)\n\n**Real-time English ↔ Mandarin speech translation for Claude — powered by Whisper, Claude AI, and Piper TTS.**\n\nGive Claude the ability to listen, translate, and speak. `live-translate-mcp` is a [Model Context Protocol (MCP)](https://modelcontextprotocol.io) server that adds speech translation as a native tool inside Claude Desktop and Claude Code. Hand it an audio file, and it transcribes, translates, synthesises, and plays the result — entirely on your machine, with Claude handling the translation.\n\n<p align=\"center\">\n  <img src=\"assets/demo.svg\" alt=\"live-translate-mcp demo\" width=\"860\" />\n</p>\n\n---\n\n## What it does\n\n| Step | Technology | Where it runs |\n|------|-----------|---------------|\n| Speech → Text | [OpenAI Whisper](https://github.com/openai/whisper) (via `@huggingface/transformers`) | Local |\n| Text → Translation | [Claude](https://anthropic.com/claude) (Opus 4.8) | Anthropic API |\n| Translation → Speech | [Piper TTS](https://github.com/rhasspy/piper) (ONNX) | Local |\n\nAudio never leaves your machine except for the translated text sent to the Claude API. ASR and TTS run fully on-device.\n\n---\n\n## Tools\n\n### `translate_file`\n\nTranslate a WAV audio file. Pass an absolute path — the server transcribes it, translates the text via Claude, synthesises speech, saves `<name>_translated.wav` next to the original, and plays it automatically.\n\n```\nTranslate /Users/alice/meeting_clip.wav\n```\n\n**Returns:** original text, translation, and the path to the saved output file.\n\n### `translate_speech`\n\nTranslate raw audio passed as a base64-encoded WAV string. Returns the transcription, translation, and synthesised audio as base64 WAV — useful for programmatic workflows.\n\n### `health_check`\n\nVerify that all dependencies (Whisper model cache, Piper voice files, `espeak-ng`) are present and ready before making a translation request.\n\n---\n\n## Installation\n\nNo installation required. Run it directly with `npx`:\n\n```bash\nnpx -y live-translate-mcp\n```\n\nOr install globally:\n\n```bash\nnpm install -g live-translate-mcp\n```\n\n**Prerequisites:**\n- Node.js 18+\n- An Anthropic API key\n- `espeak-ng` — for TTS phonemisation (`brew install espeak-ng` on macOS, `apt install espeak-ng` on Linux)\n\nThe Whisper model (~150 MB) and Piper voice models (~200 MB) download automatically on first use and are cached in `~/.live-translate/`.\n\n---\n\n## Claude Desktop setup\n\nAdd the following to your Claude Desktop configuration file:\n\n**macOS:** `~/Library/Application Support/Claude/claude_desktop_config.json`\n**Windows:** `%APPDATA%\\Claude\\claude_desktop_config.json`\n\n```json\n{\n  \"mcpServers\": {\n    \"live-translate\": {\n      \"command\": \"npx\",\n      \"args\": [\"-y\", \"live-translate-mcp\"],\n      \"env\": {\n        \"ANTHROPIC_API_KEY\": \"your-api-key-here\"\n      }\n    }\n  }\n}\n```\n\nRestart Claude Desktop. You'll see `live-translate` appear in the tools panel.\n\n---\n\n## Claude Code setup\n\n```bash\nclaude mcp add live-translate -- npx -y live-translate-mcp\n```\n\nThen set your API key in the environment before starting Claude Code, or pass it via the MCP env config.\n\n---\n\n## Usage examples\n\nOnce configured, just ask Claude naturally:\n\n> \"Translate this audio file for me: /Users/alice/recording.wav\"\n\n> \"Use translate_file on /tmp/interview.wav\"\n\n> \"Check if live-translate is ready\"\n\nClaude will call the appropriate tool automatically.\n\n---\n\n## Supported languages\n\n| Language | ASR | Translation | TTS |\n|----------|-----|-------------|-----|\n| English | ✓ | ✓ | ✓ |\n| Mandarin Chinese (普通话) | ✓ | ✓ | ✓ |\n\nLanguage is detected automatically from the audio — no need to specify it.\n\n---\n\n## How it compares\n\n| | live-translate-mcp | Cloud speech APIs |\n|--|--|--|\n| ASR | Local (Whisper) | Remote |\n| Translation | Claude API | Remote |\n| TTS | Local (Piper) | Remote |\n| Audio privacy | Audio stays on device | Audio uploaded |\n| Cost | Claude API only | Per-minute pricing |\n| Offline | Partially (ASR + TTS) | No |\n\n---\n\n## Environment variables\n\n| Variable | Default | Description |\n|----------|---------|-------------|\n| `ANTHROPIC_API_KEY` | — | **Required.** Your Anthropic API key. |\n| `PIPER_VOICE_DIR` | `~/.live-translate/voices` | Path to Piper `.onnx` voice files. |\n\n---\n\n## Related\n\n- **[live-translate](https://github.com/waxberry-dev/live-translate)** — the standalone CLI for push-to-talk translation with no API key required, using local models end-to-end.\n- [Model Context Protocol](https://modelcontextprotocol.io) — the open standard this server implements.\n- [Piper TTS](https://github.com/rhasspy/piper) — the local TTS engine powering speech synthesis.\n\n---\n\n## Licence\n\nMIT\n",
  "bytes": 4891,
  "sha": "68cc731633538996db1783c9ec1b03d02bcc36b7a75ad69749bc7959b4b2b501",
  "repo_slug": "waxberry-dev/live-translate-mcp",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_dev_waxberry_live_translate_mcp_81eeb0e9/readme"
}