Downloads · 30 days
35
39% of all-time downloads
Qwen3.6-35B-A3B-MLX-4bit
tokenfires/Qwen3.6-35B-A3B-MLX-4bit
Qwen3.6-35B-A3B-MLX-4bit is a text generation model from tokenfires. Use it when you need the model to write or continue text. It is set up for mlx. The card lists the license as apache-2.0.
4-bit MLX quantization of Qwen/Qwen3.6-35B-A3B for Apple Silicon. Quantization: affine, 4 bits, group size 64. The MoE router gates and shared-expert gates are kept at 8 bits to preserve routing accuracy.
All-time downloads
89
Public
Parameters
35.1B
20.4 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors20.4 GB · 100%
Parameter types
How the weights are stored.
U3234.7B · 99%
Try a prompt
Base models
Model card
From the Hugging Face model README
tokenfires/Qwen3.6-35B-A3B-MLX-4bit
4-bit MLX quantization of Qwen/Qwen3.6-35B-A3B for Apple Silicon. Quantization: affine, 4 bits, group size 64. The MoE router gates and shared-expert gates are kept at 8 bits to preserve routing accuracy.
Use with LM Studio
Search for tokenfires/Qwen3.6-35B-A3B-MLX-4bit in the LM Studio model downloader, or open this page and choose Use this model → LM Studio.
Use with mlx-lm
pip install mlx-lm
from mlx_lm import load, generate
model, tokenizer = load("tokenfires/Qwen3.6-35B-A3B-MLX-4bit")
messages = [{"role": "user", "content": "hello"}]
prompt = tokenizer.apply_chat_template(messages, add_generation_prompt=True)
response = generate(model, tokenizer, prompt=prompt, verbose=True)
- Task
- Text Generation
- Library
- mlx
- Type
- qwen3_5_moe
- License
- apache-2.0
- Created
- Sep 1, 2026
- Updated
- Sep 3, 2026