EXL3 quants of Qwen3.6-27B
⚠️ These are mul1 quants, intended to work with the upcoming ExLlamav3 v1.0.0 (preview in dev branch.)
They should still be compatible with earlier versions of ExLlamaV3.
2.00 bits per weight
2.50 bits per weight
3.00 bits per weight
3.50 bits per weight
4.00 bits per weight
(Benchmarks etc. to follow)
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for turboderp/Qwen3.6-27B-exl3
Base model
Qwen/Qwen3.6-27B