China's Kimi K3 AI model narrows US gap to weeks, analysts say

Beijing-based start-up Moonshot released the Kimi K3, a 2.8 trillion-parameter open-weight large language model, late last week. The model features native vision capabilities and is designed for long-horizon coding, knowledge work, and complex reasoning. Analysts say it narrows the AI gap between China and the US to weeks, challenging the conventional wisdom that Chinese models lag six to eight months behind. Ryan Fedasiuk of the American Enterprise Institute wrote that 'China has caught up in frontier AI.' Kimi K3 surpasses GPT-5.6 Sol and Claude Fable 5 on certain benchmarks, including Arena AI's front-end code development ranking, while remaining significantly cheaper than US alternatives, though pricier than other Chinese models.