Can it run?
Runs well
Using Q4_K_M in Ollama
Fit status
Runs well
Decode
398.9 tok/s
TTFT
485 ms
Safe context
95K
Memory
48.4 GB / 288.0 GB
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Agentic Coding | C | Runs well | 398.9 tok/s | 706 ms | 131K |
| Chat | C | Runs well | 398.9 tok/s | 350 ms | 50K |
| Coding | C | Runs well | 398.9 tok/s | 485 ms | 95K |
| RAG | C | Runs well | 398.9 tok/s | 882 ms | 131K |
| Reasoning | C | Runs well | 398.9 tok/s | 574 ms | 95K |
How Devstral Small 1.1 (24B params) fits at each quantization level on AMD Instinct MI350X 288GB (288.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 9.4 GB | Low | D30 |
Q3_K_S | 3 | 11.8 GB | Low | D30 |
NVFP4 | 4 | 13.4 GB | Medium | D30 |
Q4_K_M | 4 | 14.6 GB | Medium | D30 |
Q5_K_M | 5 | 17.3 GB | High | D31 |
Q6_K | 6 | 19.7 GB | High | D31 |
Q8_0 | 8 | 25.7 GB | Very High | D31 |
F16Best for your GPU | 16 | 49.2 GB | Maximum | D33 |
ollama run devstral-small-2507huggingface-cli download devstral-small-2507