Discussion

• @mistralai @mistralai on x

  🔊Introducing Voxtral TTS: our new frontier open-weight model for natural, expressive, and ultra-fast text-to-speech 🎭Realistic, emotionally expressive speech. 🌍Supports 9 languages and accurately captures diverse dialects. ⚡Very low latency for time-to-first-audio. 🔄Easily [video…

• @vllm_project @vllm_project on x

  🎉 Congrats to @MistralAI on launching Voxtral 4B TTS — enterprise-grade TTS built for production voice agents.  Day-0 support in vLLM Omni.  🌍 9 languages with natural prosody and emotional range 🎙️ 20 preset voices with easy adaptation to new ones ⚡ Ultra-low latency streaming, …

• @qtnx_ @qtnx_ on x

  Voxtral TTS it out, our first go at voice output, with really strong preference metrics! [image]

• @mistralai @mistralai on x

  Voxtral TTS is built for global applications supporting 9 languages and powering voice workflows.  ✅ Full audio intelligence: Works with Voxtral Transcribe for end-to-end speech-to-speech, or plugs into any STT + LLM stack.  ✅ Built for business: From customer support to real-tim…

• @tunguz Bojan Tunguz on x

  I just tried it out, and I am really impressed. So far my favorite AI TTS. Not as many options as some other systems, but the voice is exceptionally smooth and humanlike.

• @alex_h_liu Alexander H. Liu on x

  Both happy and sad to share my farewell project at Mistral 😭 Just amazing team and work Tech report is also available https://mistral.ai/...

• @slowdownisha Isha on x

  Mistral just dropped an open-weight TTS model: ~70ms latency ~10× real-time generation Voice AI infra is going open. [image]

• @mistralai @mistralai on x

  State-of-the-art performance. In zero-shot custom voice tests, Voxtral TTS outperformed ElevenLabs v2.5 Flash - judged by native speakers for naturalness, accent accuracy, and similarity to the original voice. [image]