Fortune said in a recent commentary that the global AI competitive landscape is changing as Chinese open-source models rise სწრაფly, with the real contest shifting from a U.S.-China national rivalry to a battle between open and closed models.
Performance and cost comparison
Citing data from independent evaluation platform Artificial Analysis, the article said DeepSeek V4 Flash trails GPT-5.6 Luna by just 1 point on its intelligence index. On pricing, it said that even after OpenAI cut prices by 80%, DeepSeek V4 Flash still carries a 60% lower per-task cost.
Export controls and the turn to open architectures
The author wrote that U.S. export control policy was intended to constrain China’s AI development, but has also had the effect of pushing Chinese companies to innovate at the architecture level and shift toward open-source, low-cost strategies. According to the article, Chinese models are attracting global developers by releasing open weights, while moving inference costs to overseas cloud service providers to support a low-cost global expansion path.
How open-source models can make money
Addressing the view that open-source models cannot generate profits, the article said their business model resembles open-source software: users pay for hosted services, and most companies do not prefer to self-deploy.
Support for open source gains traction
The piece also said that several figures in the U.S. tech sector have recently shifted toward supporting open source, including former White House AI official David Sacks. Nvidia CEO Jensen Huang has also called on the industry to embrace open source.
The author called on the United States and China to cooperate on AI safety and treat open-source AI as a “global public good,” rather than fall into what the article described as a zero-sum contest that could cost trillions of dollars.