DeepSeek Surpasses ChatGPT as Top Free App in the US
A Chinese startup, DeepSeek, has taken the U.S. tech world by surprise, surpassing ChatGPT to become the highest-rated free app on Apple’s App Store in just over two weeks.
The app, powered by the DeepSeek-V3 AI model, developed by the Hangzhou-based company DeepSeek, was released on 10 January 2025.
The meteoric rise of DeepSeek, once a hidden contender, has caught the attention of US tech users and experts as it makes waves in a market traditionally dominated by US players like OpenAI.
Rivalry with US AI Models Grows Stronger
The DeepSeek-V3 model, which developers claim "tops the leaderboard among open-source models and rivals the most advanced closed-source models globally," has raised eyebrows within the US tech community.
While DeepSeek’s technology is still a relatively unknown entity in the industry, its remarkable performance has quickly earned recognition as it competes directly with top US models, including ChatGPT-4.
Sensor Tower, an app data research firm, reported that DeepSeek has surpassed the app's rival in terms of popularity in the US, with the app’s sharp rise in downloads taking many by surprise.
DeepSeek's Rise and AI Innovation
Since its launch on January 10, DeepSeek has seen a surge in users, attracting attention for its ability to handle tasks such as math, coding, and natural language reasoning.
The startup’s DeepSeek-R1 model, released on 20 January, continues to grow in prominence, excelling in areas where ChatGPT has traditionally dominated.
According to rankings from the Chatbot Arena, DeepSeek-R1 is now tied with OpenAI’s best models in the Style Control category, and it has also secured a third-place spot overall.
Developers highlight the model’s high performance at a competitive price point, offering open-source access under an MIT license, which allows for free commercial use and further customization.
US Export Controls Under Scrutiny
The success of DeepSeek’s AI assistant has ignited discussions around the effectiveness of US export controls, which have aimed to prevent advanced American chip technologies from being used in China.
Since 2021, the Biden administration has been widening restrictions on semiconductor exports to China in an attempt to limit its technological capabilities, especially in AI.
Despite these efforts, DeepSeek’s rapid growth raises questions about the success of these measures.
Training Costs and Controversy Over Chip Use
In a report by Reuters, DeepSeek researchers revealed that the AI model was trained using Nvidia’s H800 chips, with training costs amounting to less than $6 million.
While this claim has faced challenges, it has led some US tech executives to express doubts about the efficacy of the export bans.
The fact that the chips used for DeepSeek's training are reportedly less advanced than those targeted by the US bans has added fuel to the debate.
China’s AI Boom Continues
Founded in 2023, DeepSeek is a Chinese AI lab backed by the High-Flyer hedge fund, specialising in open-source large language models.
It gained attention in 2024 after releasing the cost-effective DeepSeek-V2, triggering a price war among China's tech giants.
To date, it is the first to earn praise from the US tech industry for achieving performance levels comparable to, or even surpassing, that of leading US models.
Despite its rapid rise to prominence, little is known about the company, which continues to be an enigmatic player in the fast-evolving AI field.
As competition in the AI space grows, the battle between US and Chinese companies is far from over.