China’s DeepSeek unveils latest model a year after upending global tech | Technology News
Chinese Startup Claims New AI Chatbot Outperforms Rivals in Math and Coding
By [Author Name]
Published April 24, 2026
DeepSeek, a Chinese technology startup, has launched its latest artificial intelligence chatbot versions, DeepSeek-V4-Pro and DeepSeek-V4-Flash, positioning them as strong competitors to established U.S. models from companies like OpenAI and Google.
The new models were unveiled on Friday, roughly a year after the introduction of the company’s previous flagship, DeepSeek-R1, which garnered significant attention in the tech industry. According to DeepSeek, the V4-Pro version surpasses all competitor open models in mathematics and coding, trailing only Google’s Gemini 3.1-Pro in terms of general knowledge.
The V4-Flash model, while sharing similar reasoning capabilities with the Pro version, promises faster response times and more competitive pricing, as stated by the Hangzhou-based company.
The launch of DeepSeek-R1 in January 2025 shocked many in the tech community due to its performance, which was comparable to that of prominent models like ChatGPT and Gemini. At that time, venture capitalist Marc Andreessen referred to it as “AI’s Sputnik moment,” highlighting its potential impact on the industry.
DeepSeek’s success is notable, particularly because the company reported spending less than $6 million in computing costs—significantly lower than the multibillion-dollar investments typical in Silicon Valley.
However, the introduction of DeepSeek’s technology has not been without controversy. Several countries, including multiple U.S. states, Australia, Taiwan, South Korea, Denmark, and Italy, implemented bans or restrictions on DeepSeek-R1 shortly after its release, citing concerns over data protection and national security.
Further updates on DeepSeek’s latest models are anticipated as the story develops.
