In exactly a day and a half, Qwen 3.8 27B will absolutely DOMINATE consumer devices.
Prediction: it will have ~Opus 4.6 level performance on a single 3090, exactly as @TheAhmadOsman predicted last year.
Think about that. A local model running on a single GPU able to do things
Mike Gannotti on X: "@LLMJunky @TheAhmadOsman I was totally psyched about this until I ran comparative tests between Qwen 3.8 Max, Grok 4.5, and GLM5.2 and Qwen 3.8 wasn’t a standout…"
- I was totally psyched about this until I ran comparative tests between Qwen 3.8 Max, Grok 4.5, and GLM5.2 and Qwen 3.8 wasn’t a standout…
- Why would you expect it to beat models which are more than an order of magnitude larger?


