Chinese tech startup DeepSeek released an update to its R1 reasoning model on Thursday, drawing growing attention from multiple media outlets, with some of them saying that the latest move is set to ramp up competition with OpenAI.
According to its official WeChat account, the latest model, DeepSeek-R1-0528, is still based on the DeepSeek V3 Base model released in December 2024, but it has undergone further training with increased computational resources, significantly enhancing its depth of thinking and reasoning capabilities.
The updated R1 model has achieved top-tier performance among all domestic models across multiple benchmark evaluations—including mathematics, programming, and general logic—with overall performance approaching that of leading international models such as o3 and Gemini 2.5 Pro, according to the company.
Early Thursday, DeepSeek launched R1-0528 on the developer platform Hugging Face. International media have closely followed the latest model update by DeepSeek.
Reuters reported on Thursday that the release steps up competition with US rivals such as OpenAI. The LiveCodeBench leaderboard, a benchmark developed by researchers from UC Berkeley, MIT, and Cornell, ranked DeepSeek's updated R1 reasoning model just slightly behind OpenAI's o4 mini and o3 reasoning models on code generation, but ahead of xAI's Grok 3 mini and Alibaba's Qwen 3.
"DeepSeek earlier this year upended beliefs that US export controls were holding back China's AI advancements after the startup released AI models that were on a par or better than industry-leading models in the US at a fraction of the cost," Reuters said.
CNBC said that just as was the case with DeepSeek R1's debut, the upgraded model was released with little fanfare. It is a reasoning model, which means the AI can execute more complicated tasks through a step-by-step logical thought process, CNBC reported.
The Wall Street Journal's Chinese version reported that since the beginning of this year, DeepSeek's low-cost, high-performance R1 model has attracted global attention and sparked a surge in Chinese tech stock prices, with the market holding an optimistic view of the country's AI capabilities.
Wang Peng, an associate research fellow at the Beijing Academy of Social Sciences, told the Global Times that the widespread attention on DeepSeek's model update reflected the growing global recognition and influence of Chinese AI innovation, though facing relentless US crackdown.
"The reports of the media outlets not only affirm the country's technical strength, but also demonstrate the rising global competitiveness of Chinese AI companies, which could reshape the global AI landscape," Wang said.
In April, Chinese tech giant Alibaba released the Qwen3 model, which can switch between thinking mode for complex, multi-step tasks such as mathematics, coding, and logical deduction, and non-thinking mode for fast, general-purpose responses, according to a Xinhua report.
In March, another tech company, Baidu, released its self-developed multimodal model ERNIE 4.5, achieving collaborative optimization through joint modeling of multiple modalities, demonstrating exceptional multimodal comprehension capabilities.
Wang said that China's AI development not only drives the transformation and upgrading of its domestic economy, but also brings new opportunities for global AI technological advancement. It enables resource and achievement sharing with international partners, expands use scenarios, and collectively promotes global AI innovation and progress.
Source: People's Daily Online
