Qwen 3.8 27B: Frontier LLM that fits on your laptop

Alibaba released the open model Qwen 3.8 27B on Hugging Face, scoring 52 on the Artificial Analysis Intelligence Index, matching cloud flagship GPT-5.6 Luna. The 27B-parameter model runs locally, uses hybrid Gated DeltaNet attention, and supports reasoning depth control via the reasoning_effort parameter.

• 27B parameters, 52 AI Index points — same as GPT-5.6 Luna
• Hybrid attention: 3 of 4 layers are linear Gated DeltaNet
• Context 262,144 tokens, up to 1M with scaling
• MTP decoding in llama.cpp yields +72% tokens per second