TMTPOST – Kimi has announced the release of its next-generation math reasoning model, k0-math, with performance benchmarks comparable to OpenAI's o1 series models, specifically the publicly available o1-mini and o1-preview. Benchmark testing revealed that the k0-math model outperformed OpenAI's models in four key math assessment categories: middle school, high school, college entrance exams, and the MATH dataset, which includes beginner-level competition problems. In more challenging competition-grade datasets, OMNI-MATH and AIME, the k0-math model achieved 90% and 83% of o1-mini’s best scores, respectively, marking a significant step forward in advanced math problem-solving capabilities. Kimi founder Yang Zhilin noted that the k0-math model and the exploratory version of Kimi will gradually roll out on Kimi’s web platform and the Kimi Smart Assistant app in the coming weeks, making these advancements accessible to a broader audience. This development underscores Kimi's ambition to push the boundaries of AI-driven reasoning and compete with industry leaders in specialized domains like mathematics.