AnalysisAI ModelsAugust 21, 2026

Qwen 3.8 27B performs well even at Q3_xxs quantization

A user reports that Qwen 3.8 27B remains strong even at Q3_xxs quantization, despite typically avoiding Q3 quants due to past degradation. The user runs on an RTX 4060 Ti 16GB and notes no 35B-3AB release yet.

1 source

More stories today

Open the live feed
Qwen 3.8 27B performs well even at Q3_xxs quantization