In a bold move to challenge U.S. AI dominance, Chinese AI startup DeepSeek has released an upgraded version of its R1 reasoning model, named R1-0528. The model, now available on the Hugging Face platform, is set to escalate the competition in the artificial intelligence sector, with a particular focus on rivaling U.S. tech giants like OpenAI.
While the company did not issue an official announcement or detailed breakdown of the R1-0528’s features, initial benchmarking data from LiveCodeBench—developed by renowned universities such as UC Berkeley, MIT, and Cornell—reveals impressive results. The new version ranks just below OpenAI’s o4 mini and o3 models in terms of code generation but surpasses other models, including xAI’s Grok 3 mini and Alibaba’s Qwen 3. This places DeepSeek’s upgraded model as a formidable competitor in the increasingly crowded AI space.
The release of the R1-0528 comes on the heels of DeepSeek’s original R1 model, which disrupted the market earlier this year. Launched in January 2025, the R1 model successfully challenged assumptions that U.S. export controls would limit China’s ability to advance in AI. By offering performance comparable to U.S. models at significantly lower costs, DeepSeek forced companies like OpenAI and Google to re-evaluate their pricing and model strategies.
Industry observers had initially anticipated the launch of DeepSeek’s more advanced R2 model in May 2025. However, the launch of the R1-0528 suggests that DeepSeek is accelerating its development pace, possibly setting the stage for an even more competitive AI landscape. This latest release underscores the growing capabilities of Chinese AI startups, signalling their increasing influence on the global AI stage.
With the release of this upgraded model, DeepSeek continues to demonstrate its commitment to pushing the boundaries of AI technology. As the race between East and West intensifies, the rapid development of models like the R1-0528 could reshape the global tech ecosystem, presenting new challenges for established players.
