LLM EXPLORER 67,078 MODELS INDEXED

Qwen3.8 27B DFlash2 EXL3 4.00bpw by igor255

By igor255 · 163 downloads
HF Repo ↗ Copy URL

Qwen3.8 27B DFlash2 EXL3 4.00bpw is an open-source language model by igor255. Features: 27b LLM, VRAM: 1.2GB, Context: 256K, License: apache-2.0, LLM Explorer Score: 0.26.

  4-bit Base model:quantized:qwen/qwen...   Base model:qwen/qwen3.8-27b   Block-diffusion   Dflash2   Draft-model   Exl3   Qwen3   Region:us   Safetensors   Sglang   Speculative-decoding   Vllm
Model Card on HF 🤗: https://huggingface.co/igor255/Qwen3.8-27B-DFlash2-EXL3-4.00bpw 

Qwen3.8 27B DFlash2 EXL3 4.00bpw Parameters and Internals

LLM NameQwen3.8 27B DFlash2 EXL3 4.00bpw
Repository 🤗https://huggingface.co/igor255/Qwen3.8-27B-DFlash2-EXL3-4.00bpw 
Base Model(s)  Qwen/Qwen3.8-27B   Qwen/Qwen3.8-27B
Model Size27b
Required VRAM1.2 GB
Updated2026-09-24
Maintainerigor255
Model Typeqwen3
Model Files  1.2 GB
Model ArchitectureDFlash2DraftModel
Licenseapache-2.0
Context Length262144
Model Max Length262144
Transformers Version5.15.0
Vocabulary Size248320

Best Alternatives to Qwen3.8 27B DFlash2 EXL3 4.00bpw

Best Alternatives
Context / RAM
Downloads
Likes
Qwen3.8 27B DFlash2256K / 3.9 GB17062124
Qwen3.8 27B DFlash2256K / 3.9 GB12235148
Qwen3.8 27B DFlash2 FP8256K / 2.1 GB291
Qwen3.8 27B DFlash2 Memra256K / 3.9 GB8311
...wen3.8 27B DFlash2 EXL3 5.0bpw256K / 1.5 GB6057
...en3.8 27B DFlash2 EXL3 4.00bpw256K / 1.2 GB3882
Qwen3.8 27B DFlash2 Memra256K / 3.9 GB3292
...wen3.8 27B Heretic Ara DFlash2256K / 3.9 GB3711
...en3.6 27B Nothink DFlash2 V0.1256K / 3.9 GB1070
Qwen3.8 27B Test DFlash2 V0.3256K / 3.9 GB380
Note: green Score (e.g. "73.2") means that the model is better than igor255/Qwen3.8-27B-DFlash2-EXL3-4.00bpw.