Lotus 3.6 and Qwen 3.8 Max delivered impressive results
While the new DeepSeek V4 Flash struggled on this benchmark. Which models should I compare next? GPT-5.6 Sol Claude Opus 5 Kimi K3 Claude Fable 5 Gemini 3.6 Flash Other suggestions?
I think you can create a creative agent using OpenCode, MiniMax-H3, and Qwen 3.8 27B.
This is a physics simulator that was created purely using Qwen 3.8 27B with one prompt.
I didn't input anything other than that one prompt.
You have reached the end of the archive
All of qwen38