Downloads · 30 days
523
6% of all-time downloads
Qwen-Qwen3.8-27B-MTPLX
ben0112/Qwen-Qwen3.8-27B-MTPLX
Qwen-Qwen3.8-27B-MTPLX is a text generation model from ben0112. 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.
MTPLX-branded multi-token-prediction model for Apple Silicon (MLX). Forged with MTPLX Forge from Qwen/Qwen3.8-27B.
All-time downloads
9.2K
Public
Parameters
26.9B
16 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors16 GB · 100%
Parameter types
How the weights are stored.
U3226.9B · 100%
Try a prompt
Base models
Model card
From the Hugging Face model README
Qwen-Qwen3.8-27B-MTPLX
MTPLX-branded multi-token-prediction model for Apple Silicon (MLX).
Forged with MTPLX Forge from
Qwen/Qwen3.8-27B.
Verification
- Best depth: D3
- Multiplier vs autoregressive baseline: 2.58×
- Verified on: Apple M5 Max
- Sampler: temperature 0.6 · top_p 0.95 · top_k 20
See mtplx_runtime.json for the full verification record.
Usage
# MTPLX picks this model up automatically when downloaded:
mtplx pull ben0112/Qwen-Qwen3.8-27B-MTPLX
mtplx start chat
- Task
- Text Generation
- Library
- mlx
- Type
- qwen3_5
- License
- apache-2.0
- Created
- Aug 15, 2026
- Updated
- Aug 15, 2026