Supertonic3 Mcp
About
Local, on-device TTS for Claude & Cursor. No API key, no cloud. 10 voices, 31 languages, ~820ms on Apple Silicon.
Details
- Author
- nextic-tech
- Downloads
- 344
- Categories
- Developer Tools, Other, AI
Jump to
- No API key or cloud calls required
- 10 built-in voices: M1–M5 and F1–F5
- Supports 31 language codes including en, ko, ja, fr
- Inline expression tags: <laugh>, <pause>, <breath>, and more
- ~820ms warm synthesis speed on Apple M3 Air
- Uses FastMCP with asyncio for serialized inference
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
Supertonic3 McpCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Clone the repository, install with pip install -e ., optionally run supertonic3-mcp preload to pre-download the model, then add the server to your MCP client configuration using the absolute path to the supertonic3-mcp binary.
speak
Convert text to speech and save a WAV file. Returns the absolute path and metadata.
list_voices
List built-in Supertonic voice styles.
list_expressions
List inline expression tags (e.g. <laugh>) usable inside speak() text.
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"supertonic3 mcp": {
"supertonic3": {
"command": "uvx",
"args": [
"supertonic3-mcp"
]
}
}
}
}McpServers
{
"supertonic3": {
"command": "uvx",
"args": [
"supertonic3-mcp"
]
}
}Local, on-device TTS for Claude & Cursor. No API key, no cloud. 10 voices, 31 languages, ~820ms on Apple Silicon.
Local, on-device TTS for Claude & Cursor, powered by Supertonic 3. No API key. No cloud. An internal tool open-sourced byHalozen— we build AI compliance intelligence for construction.
Exposespeak,list_voices, andlist_expressionsto Claude Desktop, Cursor, or any MCP client overSTDIO.
git clone https://github.com/nextic-tech/supertonic3-mcp && cd supertonic3-mcp python3 -m venv .venv && source .venv/bin/activate pip install -e ".[dev]" # Optional: pre-download model for offline use (~400MB) supertonic3-mcp preload # Run MCP server (STDIO) supertonic3-mcp
Add to.cursor/mcp.json(or Cursor Settings → MCP):
{ "mcpServers": { "supertonic3": { "command": "/absolute/path/to/supertonic-tts/.venv/bin/supertonic3-mcp", "args": [] } } }
First server start downloads the Supertonic model into~/.cache/supertonic3/unless you ranpreloadfirst.
- text— 1–5000 characters; expression tags allowed
- voice_id— optional (M1,F1, …)
- language— ISO 639-1 (en,ko,ja, …).For non-English text, always setlanguage=.Defaults toen.
- speed—0.7to2.0(SDK range)
- play— iftrue, plays audio onthis machineviaafplay(macOS) oraplay(Linux). Unsupported on Windows.
WAV files are written to/tmp/supertonic_*.wav(macOS/Linux). Windows is not supported for synthesis output paths in v1.0.
Audio saved to /tmp/supertonic_abc123.wav (1.4s, voice: M1, lang: en)
Measured on Apple M3, supertonic 1.3.1 — seebenchmark/results.md.
FSL = time fromsynthesize()through WAV written (no streaming, noplay=True).
Downloads ONNX weights atomically to~/.cache/supertonic3/and prints SHA256 checksums. After preload, synthesis works without network access.
pip install -e ".[dev]" pytest
Tests mock the Supertonic SDK (no network in CI).
- listen()— Whisper speech-to-text (pip install supertonic3-mcp[stt])
- SSE transport + Docker image for remote agents
- PyPI publish workflow
MIT (this package). Supertonic SDK is MIT; model weights useOpenRAIL-M.
AI-generated speech is not a substitute for certified safety, legal, or medical guidance. For demonstration purposes only.
This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.
Interact with MiniMax's powerful APIs for text-to-speech, voice cloning, and video/image generation.
A JavaScript/TypeScript server for MiniMax MCP, offering image/video generation, text-to-speech, and voice cloning.
300+ MCP tools for macOS, all on-device — local AI inference, voice, vision, browser automation, and system control.
An AI voice toolkit with TTS, voice cloning, and video translation, now available as an MCP server for smarter agent integration.
Hosted MCP server for AudioPod's audio AI: text-to-speech, voice cloning, music generation, stem and speaker separation, transcription, denoise, and voice conversion.
AI-powered text-to-speech MCP server with instant voice cloning. Generate speech from Claude Desktop, Claude Code, or n8n using 5 built-in voices (English, German, French, Spanish) or clone any voice from a short audio sample. Runs fully local, no API keys, no cloud. Supports stdio, SSE, and HTTP transports.
An MCP server for GPT-SoVITS, providing text-to-speech synthesis, voice cloning, and multi-language support.
Local-first voice toolkit: STT (25 langs, ~19x faster than Whisper on Apple Silicon via CoreML, ONNX fallback), TTS (Kokoro + Vosk-TTS + 180 macOS voices, SSML), VAD, language detection (107 langs). Rust engine, OpenClaw skill. No cloud, no API keys.
A text-to-speech server using the macOS say command.
Production speech AI MCP server with pronunciation scoring, speech-to-text, and text-to-speech — 10 tools, 7 resources, 3 prompts.
Sign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
