Moonshot AI's Kimi K3: Advanced Reasoning & Coding AI Model

Chinese AI startup Moonshot AI has launched Kimi K3, a 2.8 trillion-parameter model designed for advanced reasoning, coding, and knowledge work. The company calls it the world’s largest open-weight AI model. This release puts Moonshot in direct competition with major U.S. AI developers. Kimi K3 includes a one-million-token context window, allowing it to process large amounts of text, images, and code in a single session. This supports longer assignments requiring sustained context. While Moonshot states Kimi K3 is behind some top models in overall comparisons, it shows stronger performance in specific coding, agentic, and hardware optimization tests. For example, independent evaluator Arena ranked Kimi K3 first in its Frontend Code benchmark. Kimi K3 uses a mixture-of-experts architecture with 896 available experts. This design aims to provide the capacity of a very large model without using every parameter for every request. Moonshot will release Kimi K3’s complete model weights on July 27th. This development highlights the increasing competition and advancements in the global AI landscape, offering new tools for developers.