Speech to textText to speech

Voxtral

Mistral's speech models: they understand audio, transcribe and answer questions about a recording. The Realtime version recognizes speech live and supports Russian; speech synthesis is also available.

Developer
Mistral AI, France
First release
Jul 2025
Latest release
Mar 2026
Sizes
3B – 24B
License
Commercial use with conditionsRecognition models (Mini 3B, Small 24B, Mini 4B Realtime) are Apache 2.0; Voxtral-4B-TTS is CC BY-NC 4.0, non-commercial only
Russian
Supported
Ready-made builds
GGUF, MLX (Apple)
Running
On your own serverAlso runs without a GPU
Industries
Customer support, Media and production, Legal
Deploy on your server Hugging Face

What it does

  • Transcribing and summarizing recordings
  • Asking questions about audio
  • Real-time recognition
  • Text-to-speech (non-commercial)

Where it is used

Call centersMediaLegal and consultingVoice assistants

Hardware requirements

LaptopLaptop or regular PC, up to 8 GB of VRAM — smaller versions
fits
1 GPUOne GPU with 16–80 GB — mid-size versions
fits
ClusterServer with several GPUs — flagship versions
no versions

Versions

  1. Voxtral 4B TTS
  2. Voxtral Mini 4B Realtime
  3. Voxtral Mini 3B и Small 24B

How to run it

On your own serverWeights are downloaded from Hugging Face and served with vLLM (load and API) or llama.cpp (modest hardware). The model then runs in a closed network with no per-request fees.Hugging Face
How much hardware you needCalculate the VRAM for the model size, context length and number of concurrent requests.Open the hardware calculator

I can set this up end to end: pick the model size, deploy it on your server and connect it to your systems. Quantization compresses a model so it takes less video memory and runs on more modest hardware. Answers change slightly, so quality is checked on your own examples.

Frequently asked questions

Can Voxtral be used in a commercial project?

With conditions. License: Recognition models (Mini 3B, Small 24B, Mini 4B Realtime) are Apache 2.0; Voxtral-4B-TTS is CC BY-NC 4.0, non-commercial only. Restrictions vary — region, company revenue, attribution requirements. Have a lawyer check the terms before a commercial launch.

What hardware does Voxtral need?

At minimum: Laptop or regular PC, up to 8 GB of VRAM — smaller versions. Some versions also run on an ordinary CPU, without a GPU. You can calculate the exact VRAM for your model size and context in the hardware calculator.

Does Voxtral support Russian?

Yes, Russian is listed on the model card.

Where can I download Voxtral and what does it cost?

The Voxtral weights are open and free to download. You only pay for the hardware it runs on and for the setup. Source links are at the bottom of this page.

How I deploy it for clients

  1. SelectionI pick the model size for your task and hardware and test it on your examples.
  2. DeploymentI deploy it on your server or in a closed network and provide an API.
  3. Fine-tuningI fine-tune it on your data (LoRA) or connect a knowledge base — whichever is cheaper for the task.
  4. IntegrationI connect it to your CRM, ERP, bot, website or team chat and set up monitoring.

Similar models

Speech to textWhisperOpenAI · USACommercial use allowed

Speech recognition in 99 languages, including Russian. The de facto standard for transcribing calls and meetings. Hugging Face's faster Distil-Whisper is English only.

DetailsSpeech to textQwen3-ASRAlibaba (Qwen) · ChinaCommercial use allowed

Speech recognition models from the Qwen team for 50+ languages, including Russian. They handle noise, singing and accents well.

DetailsTextMistralMistral AI · FranceCommercial use with conditions

European models focused on speed. Mixtral was one of the first open mixture-of-experts models; there are versions for images (Pixtral, Medium 3.5), Lean proofs and moderation (Shieldstral).

Details

Source: huggingface.co/mistralai/Voxtral-4B-TTS-2603. Data checked against the model card on 22 Sep 2026. Have a lawyer review the license before commercial launch.