Atomic releases dynamic quants for Qwen3.8 27B: 1-bit at 8.5GB, runs on 16GB MacBook Air with 92.4% token match

testingcatalog · x · 2026-08-15

Atomic has released dynamic GGUF quants for Qwen3.8-27B, compressing from 8-bit (28.9GB) to 1-bit (8.5GB). The AD-IQ3S build runs on a 16GB MacBook Air and matches the BF16 original's next token 92.4% of the time. Atomic also measured other community GGUFs, claiming closer to BF16 at most sizes.

Related event: Atomic Releases Dynamic Quantization for Qwen3.8(2 posts)→

Original post →

More from Models

Models channel →