Kimi has recently published its visual thinking model, k1, which has been developed using reinforcement learning techniques. This model is natively capable of end-to-end image understanding and thinking chains, and can be applied to a wide range of fields beyond mathematics, including physics, chemistry, and more.
In benchmark testing across various scientific disciplines, the initial k1 model outperformed leading models such as OpenAI's o1, GPT-4, and Claude 3.5 Sonnet. This achievement demonstrates the potential of k1 to revolutionize the field of artificial intelligence and its applications.
The full article can be accessed on [First Finance](https://m.yicai.com/news/102402568.html). For more updates and discussions, join us on [Telegram]([http://t.me/ZaiHuabot](http://t.me/ZaiHuabot), [Telegram Channel]([http://t.me/zaihuanews](http://t.me/zaihuanews), or [Telegram Chat]([http://t.me/zaihuachat](http://t.me/zaihuachat).