In China, Moonshot AI is preparing to release its Kimi K3 model, which boasts 2.8 trillion parameters, making it the largest open-weight AI model available for public download. The company announced that the model's weights would be released on June 27, 2026, following a successful launch earlier in the month that generated significant interest. This interest led to a sixfold increase in daily sales, prompting Moonshot to temporarily pause new subscriptions due to overwhelming demand. Users will be able to fine-tune and run the model locally for free, marking a significant step in the accessibility of advanced AI technology.
The Kimi K3 model is designed for various applications, including long-horizon coding, knowledge work, and reasoning. It has been reported to perform competitively against models from OpenAI and Anthropic, with claims that it outperforms certain versions of their models. The launch of Kimi K3 is expected to attract a large user base, particularly as Chinese AI companies continue to offer advanced capabilities at lower prices compared to their Western counterparts. This trend has been fueled by restrictions on semiconductor exports to China, which have pushed the country to develop its own chip manufacturing capabilities.
Moonshot AI's rapid growth has been notable, with the company achieving a valuation of $20 billion in May 2026 after a successful funding round led by Meituan's VC arm, Long-Z Investments. The company is now working towards a public listing within the next six months, with expectations that its valuation could exceed $30 billion. The Kimi K3 launch is anticipated to be profitable for cloud businesses that specialize in hosting third-party open models, further solidifying Moonshot's position in the competitive AI landscape.
Despite the excitement surrounding Kimi K3, the launch has also raised security concerns among Western authorities, reminiscent of reactions to previous AI model releases. As the global AI race intensifies, the implications of such advancements in technology will continue to be scrutinized, particularly regarding their potential impact on security and ethical considerations in AI deployment.
