Mistral’s Voxtral 3 TTS costs $13 per 1M characters
On October 20, Mistral released Voxtral 3, a family of speech-to-text and text-to-speech models, plus an SDK for voice agents that listen, reason, call tools and speak in a single real-time stream.
The transcription model, Voxtral Small Transcribe 3, covers 50+ languages and tells speakers apart. On FLEURS across 45 languages, Mistral reports a 4.1% word error rate with the language specified, level with Microsoft AI Transcribe 2 and well under the 16.0% of Whisper large-v3.
Voxtral Mini TTS 3 adjusts emotion, pace and emphasis to the conversation. Mistral says its $13 per 1M characters undercuts OpenAI TTS-1, Deepgram, ElevenLabs Flash and Gemini Flash. The models are available today through Mistral’s EU-hosted API or for self-deployment on your own infrastructure.
Up next
Perplexity Computer runs fully local on Macs, credit-free, Video
36 GB of unified memory is Perplexity’s recommended spec for running Computer fully local on a Mac, and in return local tasks don’t consume Computer credits. How much unified memory does your Mac have?
Related stories
- Mistral’s 1T-parameter Le Chonk keeps only 49B active
- Microsoft brings streaming MAI-Transcribe-2 to its APIs
- FLUX 3 Max outpaces Seedance 2.5 over 25x in fal’s test
- FLUX 3 Image Klein’s 5.6B count leaves out its encoder
- Nemotron 3.5 Super VL gives NVIDIA’s 120B agent model eyes
- Krea 3 edits a 2K image in under 8 seconds
Comments
No comments yet. Be the first.
Join the conversation
Sign in to leave a comment. It appears after moderation.