Mistral Small 3.2 24B Instruct at Q4_K_M needs about 16.2 GB, more than the 16 GB of the RX 7900 GRE: part of the model runs from system RAM, which slows generation. At Q3_K_M it fits entirely.
Q4_K_M · 8K context
The model needs about 16.2 GB but only 15.5 GB of VRAM is usable: 0.7 GB must be loaded into system RAM. It runs, 96% on the GPU, but more slowly.
~16.2 GB (15.6 GB–16.9 GB) needed · estimate
0.7 GB must be offloaded to system RAM
EstimateData updated 24 days agoSource: Run LLM Locally engine, Hugging Face config.json + model card
What makes this estimate less reliable
| Mistral Small 3.2 24B Instruct Q8_08K~26.7 GB · 1.7–3 tok/s estimate | Possible but slow |
|---|---|
| Mistral Small 3.2 24B Instruct Q6_K8K~21.1 GB · 3–6 tok/s estimate | CPU/GPU offload |
| Mistral Small 3.2 24B Instruct Q5_K_M8K~18.6 GB · 5–9.5 tok/s estimate | CPU/GPU offload |
| Mistral Small 3.2 24B Instruct Q4_K_M8K~16.2 GB · 13–23 tok/s estimate | CPU/GPU offload |
| Mistral Small 3.2 24B Instruct IQ4_XS8K~14.7 GB · 27–38 tok/s estimate | Compatible |
| Mistral Small 3.2 24B Instruct Q3_K_M8K~13.4 GB · 30–42 tok/s estimate | Compatible |
Prices and links follow the market you choose.
AMD
ExcellentRuns Mistral Small 3.2 24B Instruct Q4_K_M entirely in memory
Unified memory PC
ExcellentRuns Mistral Small 3.2 24B Instruct Q4_K_M entirely in memory
No current offer imported for this market: the link opens a marketplace search.
