Mistral AI · Apache 2.0
Voxtral Mini 4B Realtime
Mistral’s live transcription model for 13 languages with about half a second of delay.
How much memory it needs
Memory to run each model, from its model card or the size of its files. Smaller versions trade some quality for less memory.
| Model | Version | Needs | Business use |
|---|---|---|---|
| Voxtral Mini 4B Realtime | BF16 | 16.0 GB | Allowed |
Which computers run it
The best model in this group for each machine.
| Computer | Best version | Fit |
|---|---|---|
| Laptop or PC, no graphics card, 32 GB | Voxtral Mini 4B Realtime BF16 | Needs a graphics card |
| RTX 4060 · 8 GB | Voxtral Mini 4B Realtime BF16 | Spills into system RAM |
| RTX 3060 · 12 GB | Voxtral Mini 4B Realtime BF16 | Spills into system RAM |
| RTX 5060 Ti · 16 GB | Voxtral Mini 4B Realtime BF16 | Spills into system RAM |
| RTX 4090 · 24 GB | Voxtral Mini 4B Realtime BF16 | Runs great |
| RTX 5090 · 32 GB | Voxtral Mini 4B Realtime BF16 | Runs great |
| Mac with M4 · 16 GB | Voxtral Mini 4B Realtime BF16 | Too big |
| Mac with M4 Pro · 48 GB | Voxtral Mini 4B Realtime BF16 | Runs great |
| Mac with M5 Max · 128 GB | Voxtral Mini 4B Realtime BF16 | Runs great |
| Ryzen AI Max+ 395 · 128 GB | Voxtral Mini 4B Realtime BF16 | Runs great |
| DGX Spark · 128 GB | Voxtral Mini 4B Realtime BF16 | Runs great |
| RTX PRO 6000 · 96 GB | Voxtral Mini 4B Realtime BF16 | Runs great |
| Mac Studio with M5 Ultra · 256 GB | Voxtral Mini 4B Realtime BF16 | Runs great |
| Mac Studio with M3 Ultra · 512 GB | Voxtral Mini 4B Realtime BF16 | Runs great |
What it is good and bad at
Good at
- Live transcription
Not good at
- Needs a 16 GB graphics card
Start it
- Voxtral Mini 4B Realtime: runs in vLLM (also Transformers (>= 5.2.0), MLX (voxmlx, mlx-audio)). Model card