Text and vision retained
Quantised using oMLX v0.5.0.rc1 OQ Enhanced quantization (oQe) iMatrix
MTP Heads retained
FP16 is fastest on M1/M2 , but this can work on all MLX inferencing systems
This model is using the LATEST FROGGERIC chat template upgrade (Fixed jinja chat templates for Qwen 3.5 & 3.6 (v21))

Downloads last month: 135

Safetensors

Model size

6B params

Tensor type

F16

U32

F32

MLX

Hardware compatibility

4-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for wezzel98765/Qwen3.6-35B-A3B-oQ4e-fp16-mtp

Base model

Qwen/Qwen3.6-35B-A3B

Quantized

(757)

this model

Collection including wezzel98765/Qwen3.6-35B-A3B-oQ4e-fp16-mtp

Qwen3.6 MTP OQE FP16

Collection

8 items • Updated Jul 10