A new phase of rivalry is heating up between China and the US in the tech world. Beijing’s Moonshot Lab has introduced its Kimi K2 Thinking artificial intelligence model, which, according to developers, outperforms the latest Western AI systems—GPT-5 and Sonnet 4.5. The model is already available for free use, sparking interest from experts and entrepreneurs worldwide.
In independent tests, Kimi K2 Thinking showed outstanding skills in tackling complex analytical and academic challenges. Key benchmarks included Humanity’s Last Exam, BrowseComp, and Seal-0, where the Chinese model outperformed its competitors. The AI also handles coding tasks competently, though its edge here is less pronounced.
What sets Kimi K2 Thinking apart is its Mixture-of-Experts (MoE) architecture, which allows the system not only to plan and analyze but also to leverage external tools—such as browsers—to search for and verify information. As a result, the model can break down complicated tasks into smaller steps and find solutions even in uncertain conditions.
The developers point out that training Kimi K2 Thinking cost about $4.6 million—a sum far below the multibillion-dollar budgets of American AI labs. The model is based on the previous Kimi K2 version, released in the summer, and is now open for anyone to use, including developers and companies.
The emergence of such a tool could change the way artificial intelligence is used in business. Until now, it was believed that only paid solutions could deliver a high level of automation and analytics. Now, companies have an alternative—a free system that, in several respects, outperforms paid counterparts.
Some major American companies have already started to favor Chinese AI solutions, citing their efficiency and cost-effectiveness. However, experts warn of potential risks related to security and the origins of open-source code.
For China’s artificial intelligence industry, the launch of Kimi K2 Thinking marks a significant milestone, comparable to historic achievements in science and technology. While the US is debating the ideological aspects of the tech race, China is focusing on practicality and cost reduction.
Whether the arrival of free, powerful AI will prompt a rethink of business models in the sector remains an open question. Investors are watching developments closely, while companies worldwide are evaluating new opportunities to optimize their processes.
