LLM EXPLORER 60,147 MODELS INDEXED

Qwen3.8 Flash Next W4A16 FP8PLE by albucino

By albucino · 51 downloads
HF Repo ↗ Copy URL

Qwen3.8 Flash Next W4A16 FP8PLE is an open-source language model by albucino. Features: 72.8b LLM, VRAM: 73.7GB, License: other, LLM Explorer Score: 0.3.

  256k-context   4-bit   Auto-round   Autoround Base model:intel/qwen3.8-flash... Base model:quantized:intel/qwe...   Conversational   Endpoints compatible   Fp8   Image-text-to-text   Int4   Qwen3.8   Qwen4 exp   Region:us   Safetensors   Sharded   Speculative-decoding   Tensorflow   Vllm
Model Card on HF 🤗: https://huggingface.co/albucino/Qwen3.8-Flash-Next-W4A16-FP8PLE 

Qwen3.8 Flash Next W4A16 FP8PLE Parameters and Internals

LLM NameQwen3.8 Flash Next W4A16 FP8PLE
Repository 🤗https://huggingface.co/albucino/Qwen3.8-Flash-Next-W4A16-FP8PLE 
Base Model(s)  Qwen/Qwen3.8-Flash-Next   Intel/Qwen3.8-Flash-Next-W4A16-AutoRound   RadixArk/Qwen3.8-Flash-Next-NVFP4   Qwen/Qwen3.8-Flash-Next   Intel/Qwen3.8-Flash-Next-W4A16-AutoRound   RadixArk/Qwen3.8-Flash-Next-NVFP4
Model Size72.8b
Required VRAM73.7 GB
Updated2026-09-01
Maintaineralbucino
Model Typeqwen4_exp
Model Files  2.9 GB: 1-of-17   5.4 GB: 3-of-17   5.4 GB: 4-of-17   5.4 GB: 5-of-17   5.4 GB: 6-of-17   5.4 GB: 7-of-17   5.3 GB: 8-of-17   5.4 GB: 9-of-17   5.4 GB: 10-of-17   5.3 GB: 11-of-17   5.4 GB: 12-of-17   5.4 GB: 13-of-17   5.3 GB: 14-of-17   5.0 GB: 15-of-17   1.3 GB: 17-of-17   5.2 GB   5.2 GB   5.2 GB   5.2 GB   5.2 GB   5.2 GB   5.2 GB   5.2 GB   5.2 GB   4.4 GB
Model ArchitectureQwen4ExpForConditionalGeneration
Licenseother
Model Max Length262144
Transformers Version5.8.0.dev0
Tokenizer ClassQwen2Tokenizer
Padding Token<|endoftext|>
Errorsreplace