Qwen3.8 Flash Next GGUF & uncensored builds compared

Every popular Qwen3.8 Flash Next repo on Hugging Face in one table: quantized GGUF and MLX files, full-precision weights and derived models, including abliterated and uncensored versions. GGUF builds load in llama.cpp, Ollama and LM Studio; MLX builds run on Apple-silicon Macs. Check the memory column to see what fits your machine.

Updated 2026-11-19 · 13 repos with 1,000+ downloads in 30 days · 9 GGUF · 0 MLX · 4 other · 3 uncensored

GGUF builds

ModelQuantUncensoredLicenseDownloads (30 days)
Qwen3.8-Flash-Next-GSQ-RCO-GGUF
ISTA-DASLab
GGUF (several)–Apache 2.07.3M
Qwen3.8-Flash-Next-GSQ-RCO-abliterated-GGUF
SC117
GGUF (several)YesApache 2.04.3M
Qwen3.8-Flash-Next-GGUF
unsloth
GGUF (several)–Custom (other)3M
Qwen3.8-Flash-Next-Gyro-GGUF
agentionai
GGUF (several)–Custom (other)2.1M
Swift-1.5-Qwen3.8-Flash-Next-GSQ-RCO-GGUF
ukisai
GGUF (several)–Custom (other)1.4M
Qwen3.8-Flash-Next-GSQ-RCO-Coder-GGUF
ISTA-DASLab
GGUF (several)–Apache 2.0854.3K
Qwen3.8-Flash-Next-NVFP4-GPTQ-Strata
Maximilian228
NVFP4–Custom (other)792.7K
Qwen3.8-Flash-Next-Uncensored-GGUF
orcarouter
GGUF (several)YesApache 2.0699K
Qwen3.8-Flash-Next-Abliterated-Uncensored-Turbo-Brilliance-50x-Reasoning-Instruct-GSQ-RCO-GGUF
DavidAU
GGUF (several)YesApache 2.075.3K

Full-precision and other formats

ModelQuantUncensoredLicenseDownloads (30 days)
Qwen3.8-Flash-Next
Qwen
16-bit–Custom (other)2.9M
Qwen3.8-Flash-Next-E384-FP4-V2
autotrust
16-bit–Custom (other)62.7K
Qwen3.8-Flash-Next-FP4-Local
nvidia
16-bit–Custom (other)24.8K
Qwen3.8-Flash-Next-DFlash2
incoai
16-bit–Apache 2.03.5K

How this list is built

Hugging Face repos whose name contains “qwen3.8-flash-next”, seen in the last 30 days with at least 1,000 downloads in the last 30 days, sorted by downloads. Size comes from Hugging Face metadata, or from the repo name when the metadata has none. Memory is an estimate: parameters × bytes per weight (about 0.5 for a 4-bit build, 2 for 16-bit) plus 20% headroom; long contexts need more. “GGUF (several)” means the repo offers several quant files.

Qwen3.8 27B builds → · Qwen-Image 2.1 builds → · All uncensored models by memory → · Find a model for your hardware →