Mistral AI
STT TTS 7 models no stock voices based in France self-serve API data from API + docs synced 2026-11-16
Mistral AI is an AI lab headquartered in Paris, France, founded in April 2023. Its Voxtral audio models cover speech-to-text (batch and realtime transcription) and text-to-speech with voice cloning, served through the Mistral API on European, US and global endpoints; several Voxtral models are also published with open weights.
Speech-to-text 5
| Compare | Model | Price |
|---|---|---|
voxtral-mini-transcribe-3-0Voxtral Mini Transcribe 3Beta |
$0.21 per hour | |
voxtral-mini-transcribe-realtime-3-0Voxtral Mini Transcribe Realtime 3Beta |
$0.48 per hour | |
voxtral-small-transcribe-3-0Voxtral Small Transcribe 3Beta |
$0.24 per hour | |
voxtral-mini-2602Voxtral Mini Transcribe 2Ends 2027-04-30 |
$0.18 per hour | |
voxtral-mini-transcribe-realtime-2602Voxtral Mini Transcribe Realtime (26.02)Ends 2027-04-30 |
$0.36 per hour |
Text-to-speech 2
| Compare | Model | Price |
|---|---|---|
voxtral-mini-tts-3-0Voxtral TTS 3Beta |
$13.00 per 1M chars | |
voxtral-mini-tts-2603Voxtral TTS (26.03)Ends 2027-04-30 |
$16.00 per 1M chars |