MXFP4 Open on Hugging Face
Parameters
80B
Weights (est.)
≈ 44.8 GB
Mac memory (est.)
64 GB+
Size is estimated from the parameter count and quantisation. Memory allows a fifth extra for context plus 2 GB for the system; long contexts need more.
Run it
With mlx-lm on an Apple silicon Mac:
pip install mlx-lmmlx_lm.generate --model mlx-community/Qwen3-Coder-Next-mxfp4 --prompt "Hello"
Details
- Base model
- Qwen/Qwen3-Coder-Next (quantized)
- Vendor
- Qwen
- Task
- Text generation
- Quantisation
- MXFP4
- Licence
- apache-2.0
- Downloads
- 184 in the last 30 days · 3.6K all time · 2 likes
- Published
- 2026-05-06 · updated 2026-05-06 (6mo ago)
- Tags
- qwen3_next
Other MLX builds of Qwen/Qwen3-Coder-Next
| Build | Quantisation | Weights | Mac memory | Downloads |
|---|---|---|---|---|
| Qwen3-Coder-Next-4bit | 4-bit | ≈ 44.8 GB | 64 GB+ | 4.6K |
| Qwen3-Coder-Next-mxfp4 (this one) | MXFP4 | ≈ 44.8 GB | 64 GB+ | 184 |
| Qwen3-Coder-Next-nvfp4 | 4-bit | ≈ 44.8 GB | 64 GB+ | 116 |
| Qwen3-Coder-Next-5bit | 5-bit | ≈ 54.8 GB | 64 GB+ | 523 |
| Qwen3-Coder-Next-6bit | 6-bit | ≈ 64.7 GB | 96 GB+ | 1.8K |
| Qwen3-Coder-Next-8bit | 8-bit | ≈ 84.7 GB | 128 GB+ | 2.7K |
| Qwen3-Coder-Next-mxfp8 | MXFP8 | ≈ 84.7 GB | 128 GB+ | 142 |
| Qwen3-Coder-Next-bf16 | bf16 | ≈ 159 GB | 192 GB+ | 400 |