Params
35B / 3B act
Fastest decode
108.6 tok/s
M4 Pro 64GB · 4-bit · Rapid-MLX
Refusal
Official
Setups
3
Best reported pair · not live serving speed ↗
Runs on
Quoted one-stream decode on real boxes.
Get it running
- 1. DownloadPuts the weights on this computer.
huggingface-cli download Qwen/Qwen3.6-35B-A3B - 2. Start itUse this only if the download is GGUF.
llama-server -hf Qwen/Qwen3.6-35B-A3B - Open the weightsHugging Face ↗