Mistral Small models
3 Mistral Small models from Mistral AI on Hugging Face, from 23.6B to 24.0B parameters, published by Mistral AI in BF16, with a 32K-token context. The smallest official model, Mistral-Small-24B-Instruct-2501, needs about 52.7 GB of VRAM at its published precision; the cheapest live fit is A100 at $1.31/hr.
Part of the Mistral series · Previous generation: Mixtral · Next generation: ministral
Pick a size
One row per official Mistral Small size: the VRAM it needs at each precision, the cheapest GPU that holds it today, and the KV cache for a 32K-token context.
| Model | Parameters | Native VRAM | FP8 VRAM | INT4 VRAM | Live GPU fit (native) | Est. $/hr | KV cache at 32K |
|---|---|---|---|---|---|---|---|
| Mistral-Small-24B-Instruct-2501 | 23.6B | 52.7 GB | 26.3 GB | 13.2 GB | A100 | $1.31/hr | 5.0 GB |
VRAM is weights times a flat 1.2 overhead; the KV cache is a separate per-model figure at 16-bit, shown where the architecture is published. See the methodology.
Official models (1)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| Mistral-Small-24B-Instruct-2501 | 23.6B | BF16 | 52.7 GB | A100 | $1.31/hr |
Fine-tunes and community models (2)
| Model | Parameters | Published as | VRAM needed | Live GPU fit | Est. $/hr |
|---|---|---|---|---|---|
| Dolphin3.0-R1-Mistral-24B | 23.6B | BF16 | 52.7 GB | A100 | $1.31/hr |
| Dolphin-Mistral-24B-Venice-Edition | 24.0B | BF16 | 53.7 GB | A100 | $1.31/hr |
VRAM is for the precision the model is published in: the weights times a flat 1.2 overhead, with the KV cache not included. See the methodology. The fit shown is the lowest-priced single GPU type that holds the model at that precision, or the lowest-priced multi-GPU set (up to 8) when none does.
Run and fine-tune Mistral Small
- How to run Mistral Small: engine support, chat template and context flags, with a serving command sized per model.
- How to fine-tune Mistral Small: LoRA and QLoRA memory per model, and the license limits on derivatives.
More from Mistral AI
- Voxtral (1 models)
- Mistral 7B (7 models)
- Mixtral (1 models)
- mistralai (1 models)