LLM EXPLORER 59,962 MODELS INDEXED

Qwen3.8 Flash Next MLX 8bit MTP by Vontra

By Vontra · 2207 downloads
HF Repo ↗ Copy URL

Qwen3.8 Flash Next MLX 8bit MTP is an open-source language model by Vontra. Features: 56.6b LLM, VRAM: 202.9GB, License: other, Quantized, LLM Explorer Score: 0.34.

  8-bit   8bit   Apple-silicon Base model:quantized:qwen/qwen... Base model:qwen/qwen3.8-flash-...   Conversational   Image-text-to-text   Mixture-of-experts   Mlx   Mlx-vlm   Mtp   Omlx   Quantized   Qwen   Qwen3.8   Qwen4-exp   Qwen4 exp   Region:us   Safetensors   Sharded   Speculative-decoding   Tensorflow   Vision   Vision-language
Model Card on HF 🤗: https://huggingface.co/Vontra/Qwen3.8-Flash-Next-MLX-8bit-MTP 

Qwen3.8 Flash Next MLX 8bit MTP Parameters and Internals

LLM NameQwen3.8 Flash Next MLX 8bit MTP
Repository 🤗https://huggingface.co/Vontra/Qwen3.8-Flash-Next-MLX-8bit-MTP 
Base Model(s)  Qwen/Qwen3.8-Flash-Next   Qwen/Qwen3.8-Flash-Next
Model Size56.6b
Required VRAM202.9 GB
Updated2026-08-30
MaintainerVontra
Model Typeqwen4_exp
Model Files  4.6 GB: 1-of-42   5.1 GB: 2-of-42   5.3 GB: 3-of-42   5.0 GB: 4-of-42   5.3 GB: 5-of-42   5.0 GB: 6-of-42   5.3 GB: 7-of-42   5.0 GB: 8-of-42   5.3 GB: 9-of-42   5.0 GB: 10-of-42   5.3 GB: 11-of-42   5.0 GB: 12-of-42   4.7 GB: 13-of-42   4.9 GB: 14-of-42   4.8 GB: 15-of-42   4.9 GB: 16-of-42   4.9 GB: 17-of-42   4.8 GB: 18-of-42   4.9 GB: 19-of-42   4.9 GB: 20-of-42   4.8 GB: 21-of-42   4.9 GB: 22-of-42   4.9 GB: 23-of-42   4.8 GB: 24-of-42   4.9 GB: 25-of-42   4.9 GB: 26-of-42   4.8 GB: 27-of-42   4.9 GB: 28-of-42   4.9 GB: 29-of-42   4.8 GB: 30-of-42   4.9 GB: 31-of-42   4.9 GB: 32-of-42   4.8 GB: 33-of-42   4.9 GB: 34-of-42   4.9 GB: 35-of-42   4.8 GB: 36-of-42   4.9 GB: 37-of-42   4.9 GB: 38-of-42   4.8 GB: 39-of-42   4.9 GB: 40-of-42   4.6 GB: 41-of-42   1.0 GB: 42-of-42
Quantization Type8bit
Model ArchitectureQwen4ExpForConditionalGeneration
Licenseother
Transformers Version5.8.0.dev0