Mistral AI · First MLX build 26 Mar 2026
Voxtral-4B-TTS-2603
MLX
Audio cc-by-nc-4.0
A 4B parameter text to speech model from Mistral AI, with 3 MLX builds, from 4-bit to bf16.
- Parameters
- 4B From Hugging Face
- Quantisation
- bf16 Voxtral-4B-TTS-2603-mlx-bf16
- Memory needed
- — GB Estimate, unified memory
- Download
- — Weights, estimated
- Downloads
- 20K This build, last 30 days
- Task
- Text to speech Audio
- MLX builds
- 3 On mlx-community
- Devices
- Mac Apple silicon
Memory
Hugging Face does not list a weights size for this build, so there is no memory estimate.
Estimated from the size of the weights, plus a fifth for context and working memory and 2 GB for macOS and your apps. Long contexts need more.
Run it
Terminal · Hugging Face CLI$ pip install -U huggingface_hub
$ hf download mlx-community/Voxtral-4B-TTS-2603-mlx-bf16MLX builds
3 on mlx-community| Voxtral-4B-TTS-2603-mlx-4bit | 4-bit | 2.5 GB | 5 GB | 3.1K |
|---|---|---|---|---|
| Voxtral-4B-TTS-2603-mlx-6bit | 6-bit | 3.5 GB | 6 GB | 1.8K |
| Voxtral-4B-TTS-2603-mlx-bf16 | bf16 | — | — | 20K |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.