Qwen 3.8 27B just landed on ollama — and you can run it locally for free
No cloud login. No waiting list. And the benchmarks are where this gets interesting. What you get: → 27B parameters small enough for a capable gaming laptop or desktop → Runs offline after download →
Synth Invaders made by Qwen 3.8 27B NVFP4 running on AMD R9700 AI Pro using ROCmFPX.
ROCmFPX brings a first-class NVFP4 GGUF inference path to AMD RDNA4 and other AMD generations. ⚡ NVFP4 weights stay compressed in VRAM ⚡ GPU-side decode into an INT8-friendly representation
AI is going insane now… imagine creating this with models like; Grok 4.6 and Gemini 3.7
GTP-5.6 and Opus 5 will be costly, Kimi K3, DeepSeek V4, and Qwen 3.8 will be cheap but low quality.
8G graphics memory notebook (RTX5060) locally runs Qwen 3.8 27b speed
You have reached the end of the archive
All of qwen38