Launch more U.S. open models, issue more U.S. visas, build more U.S. data centres, and encourage more U.S. competition — these were some of the panic-tinged solutions swirling online after the Chinese AI firm Moonshot released its Kimi K3 model on July 16.
Latest Update
Also Read | Is China’s AI bet turning American AI labs into the new Apple?
Kimi K3, with 2.8 trillion parameters and a 1 million-token context window, was billed as the “world’s first open 3T-class model”. By comparison, one of Google’s recent Gemma 4 open models reported a size of 30.7 billion total parameters. Moonshot AI candidly noted that the overall performance of Kimi K3 still “trails” Anthropic’s Claude Fable 5 and OpenAI’s GPT 5.6 Sol. However, the company claimed that Kimi K3 showed frontier-level performance across its evaluation suite. On July 27, the firm enjoyed a second burst of adulation when it released the model weights, allowing users to modify and run Kimi K3 as per their commercial needs under a far more flexible licence.
Kimi K3 cost about a third of Anthropic’s Fable model in terms of the input and output token prices. While business use cases, operational needs, and other overheads influence the final cost of deploying models, such price comparisons can put makers of more expensive models on edge.
Over 270 companies, including Amazon, Google, Meta, Microsoft, Nvidia, OpenAI and SpaceX, have now signed an open letter titled ‘Open Weights and American AI Leadership’, calling for policy measures to support open-weight models, in order to keep innovation in the U.S. and ensure its technological dominance.
In September 2024, an X user asked OpenAI CEO Sam Altman when a specific AI feature would be released. Instead of a date, the asker received a mocking response. “[H]ow about a couple of weeks of gratitude for magic intelligence in the sky, and then you can have more toys soon?” posted Mr. Altman, triggering both laughter and criticism online.
But AI company-customer relations have sharply changed in 2026. Users need not wait for a “couple of weeks” and express their “gratitude” for closed model updates when Chinese AI contenders such as Alibaba, DeepSeek, and Moonshot are inviting them to try out more open models at lower prices.
The success of Kimi is not exclusively a technological triumph; it marks a win in terms of optics as well. The release of Kimi K3 — and later, its model weights — has boosted the company’s standing and its claim to “empower everyone to be superhuman”. Though not named in official missives, Kimi K3 has triggered Big Tech soul-searching in the U.S. as top companies co-sign letters of concern, express sudden support for accessible models, and brainstorm ways to race ahead of China.
As more open-weight Chinese AI models enter the arena with competitive prices and licences, companies and users will vote with their wallets. The coming weeks and months will shed more light on whether Kimi K3 was a one-shot wonder, or if Chinese firms have truly raised the bar in marketing more open AI tools.
Details
Far from uniting U.S. tech firms, however, the open letter revealed a jagged split when Anthropic did not sign. The Claude-maker instead issued a separate warning stressing the risks of open-weight models, and pushed for chip restrictions against China.
Furthermore, the AI firm backed by Amazon and Google had in February alleged that Moonshot was one of the AI labs that generated millions of exchanges with Claude through hundreds of “fraudulent” accounts in order to illicitly target the U.S. model’s agentic reasoning and tool use, coding and data analysis, computer-use agent development, computer vision, and reasoning traces.
