China's start-up Moonshot AI has claimed that its newly launched open-source model, Kimi K2, can outperform OpenAI's ChatGPT-5 and Anthropic's Claude Sonnet 4.5. Researchers behind the model said it has set "new records across benchmarks that assess reasoning, coding and agent capabilities." It reportedly beat GPT-5 on Humanity's Last Exam, a benchmark with 2,500 questions across various subjects.