Moonshot AI Releases Kimi K2.6: Open-Source Model Matches Opus 4.6 on SWE-Bench and Orchestrates 300-Agent Swarms
Chinese lab Moonshot AI has released Kimi K2.6, a one-trillion-parameter Mixture-of-Experts model that scores 58.6 on SWE-Bench Pro — surpassing GPT-5.4 and Claude Opus 4.6 — and orchestrates up to 300 sub-agents across 4,000 coordinated steps. Weights are published on Hugging Face under a Modified MIT License.


