Written by Jakub Rusinowski · Last updated November 26, 2024
Yes, but it is tight
Yes, but it is tight — OLMo 2 7B Instruct at Q6_K needs about 11.1 GB of the 12 GB usable on RTX 3060 12 GB Desktop (12 GB VRAM, 32 GB RAM), leaving only ~0.9 GB before the runtime starts swapping. Expect ~30.9 tok/s (estimated), with room for about 4,096 tokens of context.
Confidence: medium · Recommended quantization: Q6_K · Estimated speed: ~30.9 tok/s
| Usable memory for models | 12 GB |
| Memory bandwidth | 360 GB/s |
| Form factor | Desktop |
| Operating system | Windows or Linux |
| Memory upgradeable | Yes |
| Quant | Memory needed | Fits 12 GB? | Max context | Est. speed | Download |
|---|---|---|---|---|---|
| F16 | 19.7 GB | ✗ No | — | — | 14.6 GB |
| Q8_0 | 12.9 GB | ✗ No | — | — | 7.8 GB |
| Q6_K | 11.1 GB | ✓ Yes | 4K | ~30.9 tok/s | 6 GB |
| Q5_K_M | 10.3 GB | ✓ Yes | 4K | ~34.1 tok/s | 5.2 GB |
| Q4_K_M | 9.5 GB | ✓ Yes | 4K | ~37.7 tok/s | 4.4 GB |
| Q3_K_M | 8.2 GB | ✓ Yes | 4K | ~46 tok/s | 3.1 GB |
| Q2_K | 7.5 GB | ✓ Yes | 4K | ~52.4 tok/s | 2.4 GB |
| Variant | Needs at Q4_K_M | Fits? | Est. speed |
|---|---|---|---|
| OLMo 2 13B Instruct | 15.8 GB | ✗ Too large | — |
| OLMo 2 7B Instruct | 9.5 GB | ✓ Fits | ~37.7 tok/s |
Ollama or llama.cpp (CUDA/ROCm) — vLLM if you need concurrent serving
Yes, but it is tight — OLMo 2 7B Instruct at Q6_K needs about 11.1 GB of the 12 GB usable on RTX 3060 12 GB Desktop (12 GB VRAM, 32 GB RAM), leaving only ~0.9 GB before the runtime starts swapping. Expect ~30.9 tok/s (estimated), with room for about 4,096 tokens of context.
Q6_K — it needs about 11.1 GB of the 12 GB available, downloads as roughly 6 GB, and runs at an estimated 30.9 tokens/sec with up to 4K of context.
Nothing binding — the model fits with headroom and generates at a usable speed on this hardware.
Ollama or llama.cpp (CUDA/ROCm) — vLLM if you need concurrent serving