Brings Voxtral TTS voice synthesis into your MCP client with a read-only knowledge surface covering the platform's voice library, workflow patterns, and FAQ. You get two tools to list available voices and pull official links, plus resources describing the nine-language TTS setup, zero-shot cloning mechanics, and emotional style controls. No API keys required since this exposes documentation rather than live synthesis endpoints. Reach for it when you're prototyping voice features, planning multilingual audio workflows, or need Claude to answer questions about Voxtral's 70ms latency model and cross-lingual speaker consistency without leaving your editor. Cold start runs under 50ms and the whole server is stateless stdio.