SenseTime, a prominent Chinese AI company recognized for its facial recognition technology, unveiled its new open-source model, SenseNova U1, on Tuesday. The company claims that this model can generate and interpret images significantly faster than leading competitors in the United States. The introduction of SenseNova U1 marks a strategic effort for SenseTime to regain its position in China’s competitive AI landscape after experiencing setbacks amid the rising prominence of natural language processing technologies.
The innovation behind SenseNova U1 lies in its ability to analyze images directly, eliminating the need to convert them into text first. This approach not only accelerates processing time but also minimizes the computational resources required. “The model’s entire reasoning process is no longer limited to text. It can reason with images as well,” said Dahua Lin, cofounder and chief scientist at SenseTime, in an interview with WIRED.
Lin, who also serves as a professor of information engineering at the Chinese University of Hong Kong, emphasized that models capable of understanding images directly could enhance robots’ ability to interact with the physical environment. In a notable development, SenseTime announced that SenseNova U1 can be efficiently run on Chinese-made chips, addressing a significant barrier posed by US export controls that limit Chinese firms’ access to advanced AI chips mainly produced by Western companies like Nvidia. “Several Chinese domestic chipmakers have finished optimizing compatibility with our new model,” Lin stated. On launch day, ten Chinese chip manufacturers, including Cambricon and Biren Technology, confirmed their hardware support for U1.
This compatibility is critical for SenseTime, which has faced challenges in recent years due to sanctions imposed by the US government over allegations that its technology facilitated surveillance systems in Xinjiang. These sanctions have restricted US firms from investing in or selling certain technologies to SenseTime. In response, Lin acknowledged that while the company will strive to utilize a variety of domestic chips for model training, it may still rely on superior chips to maintain the speed of its iterations.
By releasing SenseNova U1 for free on platforms such as Hugging Face and GitHub, SenseTime highlights its commitment to open-source collaboration, positioning itself as a proactive player in the global AI landscape. Lin remarked that the decision to pursue an open-source strategy stemmed from the valuable feedback received from researchers, which allows for faster iterations. “In this day and age, being open source or closed source is not the winning factor; the speed of iteration is,” he explained.
SenseTime, founded in 2014, initially established itself as a world leader in computer vision technologies, including applications in facial recognition and autonomous driving. However, as conversational AI models like ChatGPT surged in popularity, the company found itself at a disadvantage, struggling to maintain profitability and falling behind nimble startups such as DeepSeek and MiniMax.
The launch of SenseNova U1 is seen as a critical effort for SenseTime to regain its footing. The company asserts that the model produces higher-quality images than existing open-source alternatives and rival technologies from leading Chinese firms, such as Alibaba’s Qwen and ByteDance’s Seedream. Nonetheless, it still trails behind industry leaders like GPT-Image-2.0, which debuted just a week before U1.
One of the standout features of SenseNova U1 is its speed, attributed to an innovative technical structure known as NEO-Unify, which SenseTime had teased earlier in the year. In an industry where rapid development and deployment are paramount, this emphasis on speed could help SenseTime regain relevance in the ongoing AI race.
As the landscape of artificial intelligence continues to evolve, SenseTime’s open-source approach may not only allow it to collaborate with international researchers but also navigate the geopolitical challenges posed by its past controversies. The company’s efforts underscore the shifting dynamics in the AI sector, where speed and flexibility are becoming critical assets in an increasingly competitive environment.
See also
Sam Altman Praises ChatGPT for Improved Em Dash Handling
AI Country Song Fails to Top Billboard Chart Amid Viral Buzz
GPT-5.1 and Claude 4.5 Sonnet Personality Showdown: A Comprehensive Test
Rethink Your Presentations with OnlyOffice: A Free PowerPoint Alternative
OpenAI Enhances ChatGPT with Em-Dash Personalization Feature
