Qwen3.8 27B Runs on Consumer GPUs as Low as 12GB VRAM

Users report running Qwen3.8 27B on consumer hardware: one achieves 18 tok/s decoding with just 12GB VRAM plus 8GB RAM, while another confirms it works on a 16GB 5060Ti using IQ3 quantization with up to 131k context.

2026-10-07 ~ 2026-10-07 · 2 related posts

• Running Qwen3.8 27B on 16GB VRAM: IQ3 quant hits 131k context at 9-30 tps — randomgenericbot · 2026-10-07
• Qwen 27B at ~18 tok/s on Just 12GB VRAM + 8GB RAM: Full Recipe Released — bodhi371 · 2026-10-07