DeepSeek, a Chinese AI platform, has made headlines by surpassing OpenAI’s ChatGPT in downloads on Apple’s App Store in the US, according to The Spectator Index.
This milestone comes just a week after DeepSeek’s official launch of its new AI model.
While established players like OpenAI, Google, and Microsoft have long dominated the space, DeepSeek’s rapid ascent highlights the growing appetite for cost-efficient and high-performing AI solutions.
DeepSeek’s technology
DeepSeek R1, developed by the Hangzhou-based startup DeepSeek, is not just another large language model—it represents a significant advancement in reasoning and analytical capabilities.
The model employs a hybrid architecture combining reinforcement learning and chain-of-thought reasoning.
Two versions are available: DeepSeek-R1 and DeepSeek-R1-Zero, the latter featuring unsupervised fine-tuning for enhanced problem-solving capabilities.
Unlike its competitors, DeepSeek has focused on making AI accessible and adaptable.
In December, DeepSeek introduced a free, open-source large language model that was said to be created in just two months with a budget under $6 million.
The lab utilised Nvidia’s H800 chips, which are less powerful than the restricted H100 chips, the latter of which are subject to US export controls.
Compact versions of the R1 model are optimised for use on laptops and other devices, ensuring flexibility beyond high-powered servers.
This versatility could be pivotal in shaping adoption trends, particularly among small businesses and individual developers.
Why is DeepSeek growing
DeepSeek’s pricing model has been a game changer. While OpenAI charges $15 per million input tokens, DeepSeek R1 costs just $0.55—a staggering 96% difference.
Despite the lower price, the model has outperformed OpenAI’s GPT-4 in specific benchmarks, achieving a 97% success rate in coding tasks.
This balance between affordability and functionality has contributed to DeepSeek’s growing popularity.
On social media platforms, users have highlighted its superior performance in tasks like mathematical computations and coding.
Its seamless integration with smartphones has further boosted user interest, catering to the demand for mobile AI applications.
Along with performance, DeepSeek’s emphasis on accessibility has positioned it as a competitor to watch.
The model is available through a user-friendly chat interface and offers developer integration via the DeepSeek Developer Portal.
By simplifying access and providing cost-effective solutions, DeepSeek is carving out a unique niche in the AI market.
Implications for the global AI market
DeepSeek’s rise underscores the intensifying competition in the AI space, particularly as Chinese tech firms gain traction on a global scale.
Its success raises important questions about how US and European companies will respond to this challenge.
While companies like OpenAI have focused on scaling capabilities, DeepSeek’s strategy of combining affordability with strong performance has set a new standard.
The development also highlights the increasing demand for AI tools that cater to a diverse user base, from casual users to professional developers.
DeepSeek’s compact versions of the R1 model, designed for use on laptops and smaller devices, exemplify this shift.
The trend towards accessible, high-performance AI tools may force other companies to rethink their pricing models and product designs.
As competition intensifies, users stand to benefit from more sophisticated, cost-effective solutions. It remains to be seen whether DeepSeek can sustain its momentum and challenge established players in the long term.
The post China’s DeepSeek outpaces ChatGPT in downloads on Apple App Store appeared first on Invezz
