Chinese AI startup DeepSeek has burst onto the global tech scene, challenging giants like OpenAI and Google with its innovative and efficient model. Founded in 2023 by Liang Wenfeng, DeepSeek has developed high-performance AI models that compete directly with industry leaders like OpenAI and Google. Its mobile app, launched in January 2025, quickly became the most downloaded free app on Apple’s US App Store, surpassing ChatGPT.
Do you want to discover the best artificial intelligence tools on the market in the field of digital marketing? Click here and download a free guide that we have prepared for you with all the information.
DeepSeek - the Chinese generative chat that is revolutionizing artificial intelligence
What is DeepSeek?
DeepSeek is an artificial intelligence startup based in Hangzhou, China. Its most prominent model, DeepSeek-R1 , has been designed to solve complex problems and has demonstrated performance comparable to the world’s best chatbots , but at a fraction of the cost. A distinctive feature of DeepSeek is that its models are open source, allowing the developer community to inspect india email data and improve the software.
What is DeepSeek
Who are its creators?
DeepSeek was founded in 2023 by Liang Wenfeng , an AI expert who previously ran the hedge fund High-Flyer, specializing in AI-based models. Liang strongly believes in the idea that innovation does not depend solely on large investments: "More investment does not necessarily lead to more innovation. If it were, large companies would have already monopolized all innovations," says Liang Wenfeng ( 36Kr ).
Liang founded DeepSeek with an initial capital of 10 million yuan and has assembled a team of talented graduates from China’s top universities. His bold and creative approach has been key to overcoming the limitations imposed by technology export restrictions to China, boosting AI development in the country.
Although both DeepSeek and ChatGPT are advanced language models, they have notable differences:
Open Source : DeepSeek-R1 is completely open source under an MIT license, allowing unrestricted commercial and academic use. In contrast, OpenAI’s ChatGPT is not completely open source.
Training efficiency : It has managed to train its models with significantly lower costs compared to OpenAI, questioning the need for large investments in advanced hardware to develop high-performance AI models.
Transparency in reasoning : It excels at showing its reasoning process before providing an answer, allowing users to better understand how a conclusion is reached. On the other hand, ChatGPT directly offers the final answer without detailing the underlying thought process.
Live Internet Access: ChatGPT currently allows real-time access to the Internet to analyze content, but DeepSeek still only offers delayed information (from the Internet until October 2023). It needs to make progress in this area to match up with tools that have already made the leap to analyzing online information in real time.