Mistral has opened the Voxtral 3 API with documented regions, capabilities and pricing; performance comparisons remain vendor-reported.
Mistral launches Voxtral 3 for production voice agents
2026-10-20Verified announcementTRIAL
Mistral has released Voxtral 3, a speech stack spanning batch and real-time transcription, text-to-speech and a voice-agent SDK. The API is available through European, US and global endpoints. Prices start at $0.0035 per minute for transcription, $0.008 per minute for real-time transcription and $13 per million characters for TTS. Mistral says the family supports more than 50 transcription languages and more than 35 TTS languages; benchmark results are vendor-run.
Availability: API available through EU, US and global endpoints; vendor also advertises self-deployment
What this means for you
For teams building voice agents, Voxtral 3 combines speech input, speech output and tool use behind one production surface with explicit regional hosting. The price is low enough for a one-day latency and accuracy trial, but vendor benchmark charts do not replace evaluation on noisy calls, accents and domain vocabulary.
Evidence & related material
Technical details & scores
| Availability | API available through EU, US and global endpoints; vendor also advertises self-deployment |
|---|---|
| Pricing | transcription from $0.0035/min; realtime transcription $0.008/min; TTS $13/1M characters |
| Language Coverage | 50+ transcription languages; 35+ TTS languages and 50+ dialects |
| Evidence Limit | accuracy, latency and TTS quality comparisons are vendor-run |
| Time Basis | official RSS and page metadata publish at 2026-10-20T12:00:22Z |
Technical impact 4
Engineering value 4
Adoption signal 2
Maturity 4
Verification cost 2
Risk 3