ISLAMABAD: Chinese artificial intelligence company DeepSeek has quietly released an upgrade to its R1 reasoning model, signaling a stronger push into the global AI race dominated by US tech firms such as OpenAI. The updated model, dubbed R1-0528, appeared on the popular developer platform Hugging Face early Thursday, though no official public announcement accompanied its release.
This silent rollout has nonetheless made a big impact. According to LiveCodeBench, a benchmark developed by leading institutions like UC Berkeley, MIT, and Cornell, the new R1 model ranks just below OpenAI’s o4 mini and o3 models in code generation performance.
Notably, it also outpaces models such as xAI’s Grok 3 mini and Alibaba’s Qwen 3, adding fuel to the notion that Chinese developers are narrowing the AI performance gap.
Quiet Release, Big Ripples in Global AI Scene
Bloomberg first reported the update, citing a message in a WeChat group from a DeepSeek representative. The message described the R1-0528 launch as a “minor trial upgrade” and encouraged users to begin testing the model.
Though minor in name, the implications are far from small. Earlier this year, DeepSeek shook global markets with the release of its R1 model, which matched or outperformed top-tier American models at a fraction of the cost. That debut sparked a wave of responses from Chinese giants like Alibaba and Tencent, who scrambled to claim AI supremacy.
The original R1 release in January was seen as a turning point, challenging assumptions that cutting-edge AI development required massive computational resources and funding.
It also spurred major US companies to adjust pricing and release smaller, more efficient versions of their models. For example, OpenAI’s o3 Mini and Google’s Gemini introduced budget-friendly tiers to compete.
Anticipation Builds for DeepSeek’s R2 Model
While DeepSeek has yet to release detailed specifications or comparisons for R1-0528, the move is being interpreted as a strategic step ahead of the expected launch of its next major model—R2. According to earlier Reuters reports, R2 was slated for a May release, though no official confirmation has emerged as of yet.
In March, DeepSeek also rolled out an upgrade to its V3 large language model, further reinforcing its position as one of China’s most formidable AI startups.
The latest R1 update suggests the company is building momentum and aiming to stay ahead in an increasingly competitive AI development landscape.
