Agentic Task: Install K3s in your own slicervm
Left: DeepSeek V4 Flash 0731 - 2x DGX Sparks Right: Qwen 3.8 27B (FP8) - 1x RTX 6000 Pro 40 (ish) tok/s gen vs 88 tok/s Why do they both feel "slow"? Thinking aka "variant" - they're spending a lot of tokens on reasoning.
Same prompt: Qwen 3.8 FP8 on the left, Qwen 3.6 FP8 on the right.
The difference in design taste isn't even close. Though it's hard to tell if 3.8 was specifically fine-tuned for this task—the level of polish is almost unreal.
Qwen 3.8 27B is insane!
Running on single 4090, with ollama opencode, prompt "generate the most fancy website you can imaging" takes ~10 min build on linux, display on iphone using Termcast port-forwarding!
Basically, this is nonsense, but it's a one-shot from Qwen 3.8 27b!
🤯 I can't believe we've reached this level of local LLMs that already deliver such a frontend on human hardware (5090). I'm in complete shock... 🚀🔥
You have reached the end of the archive
All of qwen38