Kimi K2.6 Tops OpenRouter Programming Benchmark, Outperforms Claude and GPT Series

Gate News message, April 22 — Kimi.ai announced that its latest model, Kimi K2.6, has ranked first on the OpenRouter programming ability leaderboard, becoming the leading model in developer-side evaluations.

The benchmark data shows Kimi K2.6 delivers superior comprehensive performance across programming tasks compared to mainstream models including Claude, GPT series, and other open-source models, reflecting significant improvements in code generation and development task handling. This achievement marks an important milestone for domestic AI models in real-world development scenarios, further narrowing the gap with internationally leading models.