China-based artificial intelligence startup Moonshot on July 17 unveiled Kimi K3, a 2.8 trillion-parameter open-weight model that the company describes as the world’s largest open-weight AI system. The model, built for advanced reasoning, long-horizon coding, and knowledge-intensive tasks, features a 1 million-token context window, enabling it to process far more information in a single prompt than earlier generations. Moonshot says Kimi K3 delivers performance approaching Anthropic’s frontier Fable model and reports that it performed competitively with Fable 5 while substantially outperforming OpenAI’s Opus 4.8, GPT 5.6 Sol, and GPT 5.5 in GPU kernel optimization tests. China’s Kimi K3 has also received strong third-party validation, with Arena.ai ranking it first on a benchmark for web interface-building, Vals AI placing it second overall behind Fable 5 and ahead of GPT-5.6 Sol, and Artificial Analysis finding its performance comparable to OpenAI’s GPT-5.5 and Anthropic’s Claude Opus 4.8 on complex, multi-step tasks. The launch marks the first open-weight model near the 3 trillion-parameter mark and underscores how China’s open AI ecosystem is narrowing the gap with leading US systems, following recent advances from domestic players such as Z.ai and MiniMax and amid growing global demand for powerful, customizable AI models.
Prepared by Jonathan Pierce and reviewed by editorial team.
The Kimi K3 model could change how you interact with tech. It's built for complex tasks and coding, so future apps and devices might be smarter. Keep an eye on software updates and new product launches.
China's AI game is getting stronger. The Kimi K3 model is a sign that they're catching up with US tech. Worth forwarding if you know someone interested in AI advancements.
No left-leaning sources found for this story.
No right-leaning sources found for this story.
Comments