orcarouter / Qwen3.8-27B-MLX
Agents27B1.1K27Updated Oct 27, 2026
Downloads
1.1K
Likes
27
Last updated
Oct 27, 2026
License
apache-2.0
An MLX build of the official Qwen/Qwen3.8-27B — a 27B dense hybrid-attention (Gated DeltaNet linear + full attention) native vision-language model with thinking control, tool-calling and an MTP head — quantized to MLX for Apple Silicon. Three precisions (4 / 6 / 8-bit, affine…
Tasks & library
- image text to text
- mlx
Tags
- mlx
- qwen3_5
- qwen
- qwen3.8
- apple-silicon
- quantized