LLM EXPLORER 59,962 MODELS INDEXED

Qwen3.8 Flash Next MLX 4bit MTP by macmacmacmac

By macmacmacmac · 58 downloads
HF Repo ↗ Copy URL

Qwen3.8 Flash Next MLX 4bit MTP is an open-source language model by macmacmacmac. Features: 29.3b LLM, VRAM: 102.6GB, License: other, Quantized, LLM Explorer Score: 0.29.

  4-bit   4bit   Conversational   Endpoints compatible   Image-text-to-text   Quantized   Qwen4 exp   Region:us   Safetensors   Sharded   Tensorflow
Model Card on HF 🤗: https://huggingface.co/macmacmacmac/Qwen3.8-Flash-Next-MLX-4bit-MTP 

Qwen3.8 Flash Next MLX 4bit MTP Parameters and Internals

LLM NameQwen3.8 Flash Next MLX 4bit MTP
Repository 🤗https://huggingface.co/macmacmacmac/Qwen3.8-Flash-Next-MLX-4bit-MTP 
Model Size29.3b
Required VRAM102.6 GB
Updated2026-08-30
Maintainermacmacmacmac
Model Typeqwen4_exp
Model Files  5.2 GB: 1-of-20   5.2 GB: 2-of-20   5.2 GB: 3-of-20   5.2 GB: 4-of-20   5.2 GB: 5-of-20   5.2 GB: 6-of-20   5.0 GB: 7-of-20   5.4 GB: 8-of-20   4.9 GB: 9-of-20   5.3 GB: 10-of-20   4.9 GB: 11-of-20   5.3 GB: 12-of-20   4.9 GB: 13-of-20   5.3 GB: 14-of-20   4.9 GB: 15-of-20   5.3 GB: 16-of-20   5.3 GB: 17-of-20   5.4 GB: 18-of-20   5.3 GB: 19-of-20   4.2 GB: 20-of-20   5.2 GB   0.9 GB
Quantization Type4bit
Model ArchitectureQwen4ExpForConditionalGeneration
Licenseother
Model Max Length262144
Transformers Version5.8.0.dev0
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace