Mistral AI has released Voxtral 3, a family of speech models and a voice agent SDK. The release includes three transcription models, one text-to-speech model, and an SDK integrated with LiveKit, designed for listening, reasoning, tool calls, and speaking.
Key points
- Mistral AI released Voxtral 3 speech models and a voice agent SDK.
- The family includes three transcription models and one text-to-speech model.
- Voxtral Mini Transcribe 3 is a 4B parameter model with a 32k context.
- The voice agent SDK enables real-time conversational AI applications.
- Voxtral 3 models rank highly in independent speech-to-text performance tests.
The Voxtral 3 family comprises Voxtral Small Transcribe 3, the flagship batch transcription and diarization model, which also supports translation, speech Q&A, summarization, and function calling. Voxtral Mini Transcribe 3 is a 4B parameter model for more economical batch jobs, while Voxtral Mini Transcribe Realtime 3 offers streaming ASR with end-of-turn detection.
The Voxtral Speech 3 model, also referred to as Voxtral Mini TTS 3, handles text-to-speech. All transcription models feature diarization and context biasing capabilities, with batch transcription supporting 50 languages, including 24 official EU languages. Voxtral Mini Transcribe 3 has 4B parameters and a 32k context window.
By the numbers
- Voxtral Mini Transcribe 3 parameters
- 4B
- Context window for batch models
- 32k
- AA-WER for Small Transcribe 3
- 2.1%
The Voxtral voice agent SDK operates as a three-stage streaming pipeline. It utilizes Realtime Transcribe 3 for listening and turn detection, a general LLM like Mistral Small 4 for reasoning and tool execution, and Voxtral Speech 3 for replies that match the conversation's tone. Each turn is completed in under one second.
Performance benchmarks show Voxtral 3 models ranking highly on independent speech-to-text evaluations. Small Transcribe 3 achieved a 2.1% AA-WER, placing it third on Artificial Analysis. In real-time performance, it recorded a 7.4% WER on FLEURS 35 at 240 ms, compared to Gemini 3.5 Realtime's 4.2% WER.
Sources and references
Original report
Mistral AI Releases Voxtral 3: Speech-to-Text, TTS and a Voice Agent SDK
MarkTechPost
Read on MarkTechPost (opens in a new tab)Linked in the original report
- Linkmistral.ai/news/voxtral-3 (opens in a new tab)
- Linkdocs.mistral.ai/models/voxtral-small-transcribe-3-0 (opens in a new tab)
- Linkdocs.mistral.ai/models/voxtral-mini-transcribe-3-0 (opens in a new tab)
- Linkartificialanalysis.ai/speech-to-text/non-streaming (opens in a new tab)
- Linkdocs.mistral.ai/models/voxtral-mini-transcribe-realtime… (opens in a new tab)
- Linkdocs.mistral.ai/models/voxtral-mini-tts-3-0 (opens in a new tab)
Ref 00CU5CDZ
Cite this
APA
Interestana AI Editorial. (2026, October 21). Mistral AI Launches Voxtral 3 Speech Models and SDK. Interestana. https://interestana.com/articles/mistral-ai-releases-voxtral-3-speech-to-text-tts-and-a-voice-agent-sdk-00cu5cdz
MLA
“Mistral AI Launches Voxtral 3 Speech Models and SDK.” Interestana, 21 Oct. 2026, interestana.com/articles/mistral-ai-releases-voxtral-3-speech-to-text-tts-and-a-voice-agent-sdk-00cu5cdz.
Chicago
Interestana AI Editorial. “Mistral AI Launches Voxtral 3 Speech Models and SDK.” Interestana, October 21, 2026. https://interestana.com/articles/mistral-ai-releases-voxtral-3-speech-to-text-tts-and-a-voice-agent-sdk-00cu5cdz.
BibTeX
@misc{interestana_00cu5cdz,
author = {{Interestana AI Editorial}},
title = {{Mistral AI Launches Voxtral 3 Speech Models and SDK}},
howpublished = {Interestana},
year = {2026},
month = oct,
url = {https://interestana.com/articles/mistral-ai-releases-voxtral-3-speech-to-text-tts-and-a-voice-agent-sdk-00cu5cdz},
note = {Summary of reporting by MarkTechPost, https://www.marktechpost.com/2026/10/20/mistral-voxtral-3-speech-models-voice-agent-sdk/}
}Credit the source
Original reporting: MarkTechPost, “Mistral AI Releases Voxtral 3: Speech-to-Text, TTS and a Voice Agent SDK”, https://www.marktechpost.com/2026/10/20/mistral-voxtral-3-speech-models-voice-agent-sdk/(opens in a new tab)
Follow this story
Get Artificial Intelligence in your inbox — new stories in 40 words or fewer, once a week.
Subscribe