Keep pulling the thread on Nathan Lambert.
Kimi K3 ranks #2 overall on the Vals AI index, #3 on Artificial Analysis’s Intelligence Index, and #1 in the Frontend Code Arena benchmark.
Moonshot AI is competing effectively with Anthropic and OpenAI despite possessing significantly fewer resources.
Xi Jinping publicly committed China's AI ecosystem to a future of open-source development and global diffusion at the World AI Conference.
Strong open-weight AI models significantly reduce the potential profit margins for closed-source AI labs.
The architectural and data recipe changes in Kimi K3 resulted in an approximate 2.5x improvement in overall scaling efficiency compared to Kimi K2.
Chinese AI labs are significantly more capital-efficient than their American counterparts.
The release of Kimi K3 increases the probability that China could lead in AI capabilities in the near future due to more efficient training methods.
Alibaba has announced it will soon release Qwen 3.8, a 2.4 trillion parameter model, with open weights.
Moonshot AI released its flagship model Kimi K3 on July 16th, 2026.
Moonshot AI's Kimi K3 is a 2.8 trillion parameter Mixture of Experts (MoE) model.
Moonshot AI plans to release the weights for its Kimi K3 model on July 27th, 2026.
The performance gap between the best open-weight and closed-weight AI models has narrowed from a debated 6-9 months to approximately 3-5 months.