# Any uncensored qwen 3.8 flash next GGUF with full BF16 N-gram? — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/c5786ee48ae6d06d

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-10-20 15:59 ingested 2026-10-21 02:41

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1xavzra/any_uncensored_qwen_38_flash_next_gguf_with_full/)

## retained text



```
Im fed up with qwen 3.8 flash next refusing to do certain tasks (e.g. finding vulnerabilites in my code or scanning network related stuff). 

Right now Im using [https://huggingface.co/SassyDiffusion/Qwen3.8-Flash-Next-GGUF/tree/main](https://huggingface.co/SassyDiffusion/Qwen3.8-Flash-Next-GGUF/tree/main) which is an unsloth quant with full precision N-gram. 
```



document fetched 2026-10-21 00:41 · canonical https://reddit.com/r/LocalLLaMA/comments/1xavzra/any_uncensored_qwen_38_flash_next_gguf_with_full
