DeepSeek Tops the Free App Charts in Both the US and China Apple Stores
The large model storm stirred up by DeepSeek is escalating in both China and the United States.
On January 27th, the Apple App Store’s free list in China showed that DeepSeek had reached the top spot. At the same time, DeepSeek soared from sixth place to first place on the free list of Apple App Store in the US, surpassing ChatGPT, social media platform Threads under Meta company, as well as generative AI products from American tech companies like Google Gemini and Microsoft Copilot.
This is also the only time a Chinese app has been able to simultaneously occupy the top spot in both Chinese and US Apple App Stores. Previously, Xiaohongshu and Temu have also become champions of downloads in the US App Store but failed to achieve simultaneous top rankings in China.
Just yesterday, DeepSeek experienced a short-term flash crash. In response to this, DeepSeek stated that there was indeed a local service fluctuation in the afternoon, but the issue was resolved within minutes. This incident may have been caused by a surge in user traffic after the release of a new model, leading to servers temporarily unable to meet the high concurrency demands of users.
The application is developed by the domestic large-scale model company ‘DeepSeek’ (affiliated with the quantitative giant High-Flyer Quant). Its inference large model DeepSeek-R1 quickly gained market attention after its release due to its high performance and low cost advantages.
As an open-source model, DeepSeek-R1 performs close to OpenAI’s GPT-4o in mathematical, code, and natural language reasoning tasks, with training costs only one-tenth of the latter (approximately $5.576 million). Its reinforcement learning-driven training method is considered innovative and has received high praise from NVIDIA scientist Jim Fan.
In the overseas developer community, this model is highly praised for its support of free commercial use and derivative development, and ranks among the top in evaluations such as Chatbot Arena.
The simultaneous topping of DeepSeek in both the US and China Apple App Stores signifies a significant breakthrough for Chinese AI applications in the international market. This also confirms the evaluation of DeepSeek’s recent performance by Zhou Hongyi, founder of 360 company – ‘simply amazing’.
This event holds important significance for both DeepSeek itself and the industry. It not only marks the first time that Chinese AI product downloads have surpassed leading American products like ChatGPT in mainstream overseas markets but also challenges the traditional model where large companies rely on massive capital investment. Furthermore, DeepSeek’s success could boost confidence in domestically developed large models within China or accelerate capital flow towards domestic AI technology. Its open-source strategy (such as releasing 6 small models) is expected to promote collaboration within China’s AI developer community.
Microsoft CEO Nadella and Meta engineers mentioned on anonymous forums that DeepSeek’s low-cost high-efficiency model has forced American companies to reevaluate their research and development strategies. Meta was reported to be ‘reverse engineering’ its technology urgently. This may accelerate American companies shifting from closed-source to more open collaborative models.
However, at the moment of reaching the top spot, DeepSeek faces multiple challenges related to geopolitics, technological paths, and business sustainability. Once user novelty fades away, whether DeepSeek can continue consolidating its technological advantage to withstand competition pressures from all sides will determine its next steps.
