Mistral Small 24B hardware requirements
23.6B parameters. Needs about 16.2 GB at Q4_K_M with 8K context. Model card.
Memory needed at each quantization
| Quant | Quality | 8K context | 32K context |
|---|---|---|---|
| FP16 | Full quality | 49.1 GB | 53.2 GB |
| Q8_0 | Near-lossless | 27 GB | 31 GB |
| Q6_K | Excellent | 21.3 GB | 25.3 GB |
| Q5_K_M | Very good | 18.7 GB | 22.8 GB |
| Q4_K_M | Good, most popular | 16.2 GB | 20.3 GB |
| Q3_K_M | Noticeably degraded | 13.5 GB | 17.5 GB |
| Q2_K | Heavily degraded | 10.8 GB | 14.8 GB |
Run it with Ollama: ollama run mistral-small:24b (downloads Q4_K_M by default).
Run it in the cloud
Don't want to buy hardware? Run Mistral Small 24B on a rented GPU today.
Rent a cloud GPU by the hour. No upfront cost, cancel any time.
Every GPU and Mac checked
| GPU | Runs? | Best fit |
|---|---|---|
| NVIDIA RTX 3060 12GB | Yes, with heavy quantization | Q2_K · ~10.8 GB · ~24 tok/s |
| NVIDIA RTX 3070 | No | Needs ~16.2 GB at Q4_K_M; 8 GB available |
| NVIDIA RTX 3080 10GB | No | Needs ~16.2 GB at Q4_K_M; 10 GB available |
| NVIDIA RTX 3090 | Yes | Q6_K · ~21.3 GB · ~29 tok/s |
| NVIDIA RTX 4060 | No | Needs ~16.2 GB at Q4_K_M; 8 GB available |
| NVIDIA RTX 4060 Ti 16GB | Yes, with heavy quantization | Q3_K_M · ~13.5 GB · ~15 tok/s |
| NVIDIA RTX 5060 Ti 16GB | Yes, with heavy quantization | Q3_K_M · ~13.5 GB · ~23 tok/s |
| NVIDIA RTX 4070 | Yes, with heavy quantization | Q2_K · ~10.8 GB · ~34 tok/s |
| NVIDIA RTX 4070 Ti Super | Yes, with heavy quantization | Q3_K_M · ~13.5 GB · ~35 tok/s |
| NVIDIA RTX 4080 | Yes, with heavy quantization | Q3_K_M · ~13.5 GB · ~37 tok/s |
| NVIDIA RTX 4090 | Yes | Q6_K · ~21.3 GB · ~31 tok/s |
| NVIDIA RTX 5070 | Yes, with heavy quantization | Q2_K · ~10.8 GB · ~46 tok/s |
| NVIDIA RTX 5070 Ti | Yes, with heavy quantization | Q3_K_M · ~13.5 GB · ~47 tok/s |
| NVIDIA RTX 5080 | Yes, with heavy quantization | Q3_K_M · ~13.5 GB · ~50 tok/s |
| NVIDIA RTX 5090 | Yes | Q8_0 · ~27 GB · ~43 tok/s |
| AMD Radeon RX 7800 XT | Yes, with heavy quantization | Q3_K_M · ~13.5 GB · ~32 tok/s |
| AMD Radeon RX 7900 XTX | Yes | Q6_K · ~21.3 GB · ~30 tok/s |
| Apple M4 (16GB) | No | Needs ~16.2 GB at Q4_K_M; 10.7 GB available |
| Apple M1 (16GB) | No | Needs ~16.2 GB at Q4_K_M; 10.7 GB available |
| Apple M2 Pro (32GB) | Yes | Q6_K · ~21.3 GB · ~6 tok/s |
| Apple M3 Max (64GB) | Yes | Q8_0 · ~27 GB · ~10 tok/s |
| Apple M4 Pro (48GB) | Yes | Q8_0 · ~27 GB · ~7 tok/s |
| Apple M4 Max (128GB) | Yes | FP16 · ~49.1 GB · ~7 tok/s |
| Apple M2 Ultra (192GB) | Yes | FP16 · ~49.1 GB · ~10 tok/s |