Voice

Mistral Voxtral 3

Speech-to-text and text-to-speech models plus a voice-agent SDK, via API or self-hosting.

Mistral's new speech-to-text and text-to-speech model family (50+ languages) plus a voice-agent SDK for real-time phone/web voice agents, via EU-hosted API or self-deployment.

Vendor: Mistral AI
Price: Paid
Experience: Advanced
Platforms: API
Last reviewed: October 22, 2026
Free tier: No

Visit Mistral Voxtral 3

This button opens Mistral AI’s site. The address is the official URL in our catalog, not an affiliate link. Disclosures.

Our rating

Overall rating: rank and stars across the whole category. A match % on guides and results is for one job and your answers, so a tool can be a top match and still rank lower overall.

Tied · #6 of 10 in Voice

Quality: 4
Features: 4
Value: 3
Ease: 2

3.5 out of 5

regular price not published; value neutral

Mistral's speech-to-text and text-to-speech models plus a voice-agent API, sold by usage for developers.

Reviewed October 22, 2026.

Pros and cons

Pros

Text-to-speech costs $13 per 1M characters
Batch transcription from $0.0035 a minute
Voice-agent stack works in 50+ languages and can run on-premises

Cons

API only, with no consumer app
No free tier listed on the launch page

Jobs it fits

Create a voiceover — Speech-to-text and text-to-speech models plus a voice-agent SDK, via API or self-hosting. Fit score 70.