A Chinese artificial intelligence company just released a model that matches the best from the US - and the stock market did not take it well.
Yang Zhilin, a prior Meta and Google employee who also served as a Tsinghua University professor, launched Moonshot AI in early 2023. The company built Kimi, a chatbot that became China's most popular alternative to ChatGPT, and hit a $20 billion valuation along the way. According to a June Bloomberg report, the startup had annual recurring revenue exceeding $200 million as of April 2026 and now seeks roughly $2 billion in fresh capital at a $30 billion valuation. The company is reorganizing its offshore holdings to facilitate a Hong Kong initial public offering.
Now it has released Kimi K3, a model that the company acknowledges is not perfect. Moonshot said in its release, "despite being a highly competitive model overall, K3 nonetheless exhibits a noticeable gap in user experience compared with Claude Fable 5 and GPT 5.6 Sol."
Get the market news that matters in a five-minute read with Market Briefs, our free daily newsletter
According to Artificial Analysis, a benchmarking site, the average expense for a standardized intelligence task on Kimi 2.6 or DeepSeek's V4 Flash falls between 2 and 33 cents. For the same task, Anthropic's Claude Fable 5 costs $2.75.
That is roughly a third of the price of the closest US competitor. For companies that pay for AI by the query, that difference adds up fast.
The stock reaction came down to one fear: if US companies can get nearly the same performance from a Chinese model at a fraction of the cost, they might not need to keep buying as many expensive Nvidia chips or building as many massive data centers. Nvidia, SK Hynix, and Samsung all saw their stock prices drop after the Kimi K3 announcement.
There is also bad blood between the companies. Anthropic has accused Moonshot, along with DeepSeek and MiniMax, of running "industrial-scale campaigns" that it says are "in violation of our terms of service and regional access." This practice, known as distillation, involves employing a larger, costlier model to instruct a more compact, budget-friendly version, often done without authorization. Anthropic detailed these allegations in a blog post this past February. Moonshot did not reply when asked for comment.
The bottom line: Should Kimi K3 demonstrate strong performance in coding - a highly profitable segment that generates significant revenue for Anthropic and OpenAI before their planned IPOs - it could undermine the business model of those competitors.
The firm's growth trajectory underscores the rapidly rising profile of Chinese AI startups in a market long dominated by US giants. The release of Kimi K3 is a reminder that competition from China is real, and it is getting cheaper by the month.
Join Market Briefs, our free daily newsletter, for a quick daily rundown of the markets
