Project Overview:

Kimi released the K1 Visual Thinking Model, built on reinforcement learning technology. It natively supports end-to-end image understanding and chain-of-thought techniques, extending its capabilities beyond mathematics to a wider range of fundamental scientific fields. In benchmark capability tests across core disciplines such as mathematics, physics, and chemistry, the first-generation K1 model outperformed global benchmark models like OpenAI's O1, GPT-4O, and Claude 3.5 Sonnet.

​

Kimi's groundbreaking K1 Visual Thinking Model is here! A powerful combination of exceptional vision and advanced reasoning – see what you observe, solve what puzzles you.