Nvidia is pouring $6 billion into creating an open AI model to counter China's DeepSeek and Kimi K3.
Nvidia is betting on open source in its fight for dominance in the field of artificial intelligence. My analysis of the situation shows that the $6 billion deal with the AI startup Poolside is not just an investment, but a strategic maneuver aimed at creating a powerful alternative to Chinese large language models, including DeepSeek and Kimi K3 from Moonshot AI.
Agreement details and deal structure
According to my data, Nvidia has agreed to pay Poolside $6 billion for licenses to AI models, as well as offer jobs to more than 100 startup employees. In addition, the company is ready to invest an additional $1 billion in the project's development. This structure resembles the recent acquisition of chip developer Groq, where Nvidia paid $20 billion for licenses and integrated most of the employees and founders into its team.
This move underscores Nvidia's strategy of supporting models with open weights, which developers can freely use and modify. The company seeks to democratize access to advanced technologies and prevent the concentration of AI power in the hands of a few giants. At the same time, it is a direct signal to OpenAI and Anthropic: open models are cheaper to operate and easier to customize, making them serious competitors to closed ecosystems.
Political context and market position
Notably, the deal comes amid renewed discussions in the U.S. presidential administration about possible restrictions on the use of Chinese open-weight models. In July, 25 American companies, including Nvidia, Meta, and Microsoft, issued an open letter warning that bans would not strengthen U.S. technological leadership but would merely shift the market toward closed developers. The escalation of these debates was largely triggered by the release of Kimi K3, whose creators the White House accused of distilling Anthropic's Fable model.
In my view, Nvidia is acting proactively here, trying to kill two birds with one stone: on the one hand, to satisfy demand for its chips through the growth in the number of open models, and on the other, to create a tool that will reduce the American market's dependence on Chinese developments. The only question is whether Poolside, with $6 billion and Nvidia's support, can truly compete with giants like DeepSeek—but the bet is clearly on openness and speed of adoption.