The DeepSeek app for iOS has surged ahead of OpenAI’s ChatGPT, claiming the top spot on the App Store’s “Top Free Apps” chart in the United States. Released by the Chinese company DeepSeek, the open-source DeepSeek-R1 artificial intelligence (AI) model has surpassed OpenAI’s o1 AI model in various benchmarks. This development has generated considerable conversation within tech circles, particularly among Silicon Valley leaders who are reacting to the app’s rapid ascent. Importantly, the DeepSeek app is free, with no subscription tiers introduced to date.
DeepSeek Overtakes ChatGPT
Since the launch of the ChatGPT app for iOS in May 2023, it has consistently ranked among the top free applications on the App Store and was previously the most prominent AI app available. However, DeepSeek recently claimed the top position on the chart following the debut of the DeepSeek-R1 AI model.
Despite being relatively obscure, the Hangzhou-based AI company was established in 2023 and has developed several open-source large language models (LLMs). Although there are other US tech firms releasing open-source models, such as Meta, none have matched the scale or performance of DeepSeek’s offerings. The company has also stated that the development cost of the DeepSeek-R1 model was $6 million (approximately Rs. 51.8 crore), a figure significantly lower than typical costs for models of this complexity.
The rapid rise of DeepSeek’s technology has caught the attention of several prominent figures in Silicon Valley. Venture capitalist Marc Andreessen described the R1 AI model as “one of the most amazing and impressive breakthroughs I’ve ever seen.” Similarly, Aravind Srinivas, the Co-Founder and CEO of Perplexity AI, praised DeepSeek for being the first AI application to surpass ChatGPT in the App Store rankings.
Yann LeCun, a pioneering figure in AI and the current Vice President and Chief AI Scientist at Meta, expressed on LinkedIn that “open source models are surpassing proprietary ones.” Holger Zschäpitz, a senior financial reporter at Germany’s Welt, also remarked that DeepSeek “could represent a big threat to US equity markets.”
DeepSeek Is Not the Only One
While DeepSeek has shifted the paradigm by open-sourcing advanced AI models rather than offering only limited capabilities, it is not the sole player in this arena. Another Chinese company, Kimi AI, has disclosed the launch of its Kimi k1.5 AI model, claiming it to be an “o1-level multimodal LLM” that outperforms both GPT-4o and Claude Sonnet 3.5 in several evaluations.
Kimi AI has also made its web version of the chatbot freely accessible with unlimited usage, akin to DeepSeek’s offering. The Kimi k1.5 model boasts capabilities such as conducting real-time web searches, analyzing up to 50 files in various formats, and understanding images. While its technical report can be found on GitHub, the model itself is not open-source at this time.
