I know it's hard to believe, but this IQ_3 of Qwen 3.8 Flash Next is genuinely impressive.
Sure, it's not BF16, but it works very well in Claude Code. To eliminate the likelihood of benchmaxxing, I present a novel voxel benchmark: Voxel Sermon on the Mount. This was generated
GPT-6.1 Sol doesn’t need to write the code.
It can just be the brain. Sol plans + reviews. Qwen 3.8 27B writes locally. $0.75 → $0.17 API cost for 3 small games. That’s a 77% cut. The trade-off? 6.6 min → 43.4 min. Cloud brain. Local hands.
Gpt-6 luna, qwen 3.8 and deepseek v4 are FREE right now, and none of these are in the article yet
Nearest cutoff first: Nous Portal: Step 3.7 Flash until Oct 1, plus LongCat 2.5 Preview, Solar Pro 4, Laguna S 2.1 on the $0 plan: Qoder: Qwen3.8-Flash
I have been grinding the Omarchy Local AI from 0xSero and loving it.
I installed on my Omarchy Franken GPU (3x 3090) and had Qwen 3.8 27B running on 2 and Qwen Image 2.1 running on another. With the integrated Pi I quickly created: - Local Image generation engine and website -
You have reached the end of the archive
All of qwen38