On Monday, Chinese startup DeepSeek’s AI Assistant surpassed ChatGPT to become the top-rated free app on Apple’s App Store in the United States. Powered by the DeepSeek-V3 model, which its creators claim “leads the open-source model rankings and competes with the most advanced closed-source models globally,” the app has gained significant popularity among U.S. users since its release on January 10, according to app data research firm Sensor Tower.
This milestone underscores how DeepSeek has made a significant impact in Silicon Valley, challenging the long-held belief in U.S. dominance in AI and raising questions about the effectiveness of Washington’s export controls aimed at limiting China’s access to advanced chips and AI technologies.
AI models, from ChatGPT to DeepSeek, require powerful chips for training, and the Biden administration has expanded restrictions since 2021 to prevent these chips from being exported to China for use in training AI models.
However, DeepSeek’s researchers revealed last month that they used Nvidia’s H800 chips, costing less than $6 million for training, to build the DeepSeek-V3. While this claim has been contested, the assertion that the chips were less powerful than Nvidia’s top-tier models that the U.S. aims to keep out of China, combined with the relatively low training costs, has prompted U.S. tech leaders to question the effectiveness of these export controls.
Despite the growing attention, much remains unknown about DeepSeek, a small startup based in Hangzhou, founded in 2023 following Baidu’s release of China’s first AI large-language model. Since then, several Chinese companies have rolled out their own AI models, but DeepSeek is the first to earn praise from U.S. tech industry experts for potentially matching or even exceeding the performance of leading U.S. models.
