Chinese Startup Moonshot AI Unveils Kimi K3 AI Model

Chinese AI firm Moonshot AI has unveiled its Kimi K3 model, stating it closes the performance gap with leading U.S. AI systems. While Kimi K3 still trails Anthropic's Claude Fable 5 and OpenAI's GPT 5.6 Sol in overall performance, the company reported it consistently outperformed other tested models.

According to Moonshot AI, Kimi K3 beat Claude Opus 4.8 and GPT 5.5—models just below Anthropic and OpenAI’s cutting-edge systems—on benchmarks including coding and general agents. The model is China’s largest to date with 2.8 trillion parameters.

Bank of America analysts noted that K3 demonstrates that pre-training scaling and architectural innovation can still yield significant gains for flagship Chinese models, despite hardware constraints in China. This development occurs as the race for AI supremacy between the U.S. and China intensifies.

Chinese AI models are gaining traction among Western companies due to their competitive pricing and rapidly improving performance. Following the announcement, shares of Chinese AI rivals Z.ai and MiniMax Group dropped by 28% and 16%, respectively.

Founded in 2023 in Beijing, Moonshot AI is a prominent player in China's AI development landscape. The release of Kimi K3 marks a step forward in the capabilities of Chinese AI models.