Voxtral Mini TTS 3 promises expressive speech in 35-plus languages

Mistral says Voxtral Mini TTS 3 achieves state-of-the-art expressiveness in 35-plus languages and 50-plus dialects. It says the model reads emotion, pace and emphasis from conversation or follows line-by-line tags for emotion and pace. The company is also rolling out Pro Voice, which it says can replicate a voice across supported languages from up to 30 minutes of recorded audio.