On the Humanity’s Last Exam (HLE) benchmark, Kimi K2.5 scored 50.2% (with tools), surpassing OpenAI’s GPT-5.2 (xhigh) and Claude Opus 4.5. It also achieved 76.8% on SWE-bench Verified, cementing its ...
At this year's Build conference, Microsoft unveiled a major expansion of its agent-based AI platform, highlighting new tools to securely build, customize and orchestrate intelligent agents across ...