Olmo 3.1 32B Instruct at Q4_K_M needs about 21.1 GB, more than the 12 GB of the RTX 3060 12 GB: part of the model runs from system RAM, which slows generation. At Q3_K_M it does not fit in VRAM.
Q4_K_M · 8K context
The model needs about 21.1 GB but only 11.6 GB of VRAM is usable: 9.5 GB must be loaded into system RAM. It runs, 55% on the GPU, but more slowly.
Estimated speed is at most 3.5 tokens/s: usable, but slow.
~21.1 GB (20.4 GB–21.8 GB) needed · estimate
9.5 GB must be offloaded to system RAM
EstimateData updated 40 days agoSource: Run LLM Locally engine, Hugging Face config.json + model card
What makes this estimate less reliable
| Olmo 3.1 32B Instruct Q8_08K~35.6 GB · 0.8–1.6 tok/s estimate | Not recommended |
|---|---|
| Olmo 3.1 32B Instruct Q6_K8K~27.9 GB · 1.2–2.5 tok/s estimate | Possible but slow |
| Olmo 3.1 32B Instruct Q5_K_M8K~24.4 GB · 1.5–3 tok/s estimate | Possible but slow |
| Olmo 3.1 32B Instruct Q4_K_M8K~21.1 GB · 2–3.5 tok/s estimate | Possible but slow |
| Olmo 3.1 32B Instruct IQ4_XS8K~19 GB · 2.5–4.5 tok/s estimate | Possible but slow |
| Olmo 3.1 32B Instruct Q3_K_M8K~17.3 GB · 3–6 tok/s estimate | CPU/GPU offload |
Prices and links follow the market you choose.
AMD
CompatibleRuns Olmo 3.1 32B Instruct Q4_K_M entirely in memory
Apple Silicon
CompatibleRuns Olmo 3.1 32B Instruct Q4_K_M entirely in memory
Unified memory PC
CompatibleRuns Olmo 3.1 32B Instruct Q4_K_M entirely in memory
No current offer imported for this market: the link opens a marketplace search.
