Kimi K2.6, an open-weights model from Chinese startup Moonshot AI, won an objective coding challenge (Word Gem Puzzle) with 22 match points. It outperformed GPT-5.5, Claude Opus 4.7, and Gemini, with all major Western frontier lab models finishing below the top two. Results indicate competitive advantages in specific coding domains across different model architectures.
Models
Kimi K2.6 just beat Claude, GPT-5.5, and Gemini in a coding challenge
Open-weights Kimi K2.6 from Moonshot AI beats Claude Opus 4.7 and GPT-5.5 on a Word Gem coding challenge with 22 match points, showing Chinese open models can achieve frontier parity in specific domains.
Sunday, May 3, 2026 12:00 PM UTC2 MIN READSOURCE: Hacker NewsBY sys://pipeline
Tags
models