# Qwen 3.8 Flash Next 'Flare' images for 1M context — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/52ad70b745a1d2b7

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-10-30 19:28 ingested 2026-10-31 02:40

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1xjowbm/qwen_38_flash_next_flare_images_for_1m_context/)

## retained text



```
For those using RTX 6000 PRO cards, I've published two images on huggingface:

  

[https://huggingface.co/JohannesBertens/Qwen3.8-Flash-Next-FLARE-1M](https://huggingface.co/JohannesBertens/Qwen3.8-Flash-Next-FLARE-1M)

[https://huggingface.co/JohannesBertens/Qwen3.8-Flash-Next-FLARE-1M-TP2](https://huggingface.co/JohannesBertens/Qwen3.8-Flash-Next-FLARE-1M-TP2)

  

The first is for 1 GPU, the second for 2 GPUs. I like the flare inference engine, I like Qwen 3.8 Flash Next, but I also need 1M context for my agents. This was the solution.

  

Hope it helps!
```



document fetched 2026-10-31 00:44 · canonical https://reddit.com/r/LocalLLaMA/comments/1xjowbm/qwen_38_flash_next_flare_images_for_1m_context
