Qwen3.8 27B at Q4_K_M needs about 18.2 GB, more than the 8 GB of the RTX 4060: 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 18.2 GB but only 7.7 GB of VRAM is usable: 10.5 GB must be loaded into system RAM. It runs, 42% on the GPU, but more slowly.
~18.2 GB (17.4 GB–18.9 GB) needed · estimate
10.5 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
| Qwen3.8 27B Q8_08K~30.1 GB · 0.9–1.7 tok/s estimate | Not recommended |
|---|---|
| Qwen3.8 27B Q6_K8K~23.7 GB · 1.2–2.5 tok/s estimate | Possible but slow |
| Qwen3.8 27B Q5_K_M8K~21.1 GB · 1.4–3 tok/s estimate | Possible but slow |
| Qwen3.8 27B Q4_K_M8K~18.2 GB · 1.8–3.5 tok/s estimate | Possible but slow |
| Qwen3.8 27B IQ4_XS8K~16.9 GB · 2–4 tok/s estimate | Possible but slow |
| Qwen3.8 27B Q3_K_M8K~15.3 GB · 2.5–4.5 tok/s estimate | Possible but slow |
Prices and links follow the market you choose.
AMD
CompatibleRuns Qwen3.8 27B Q4_K_M entirely in memory
Unified memory PC
CompatibleRuns Qwen3.8 27B Q4_K_M entirely in memory
No current offer imported for this market: the link opens a marketplace search.
