SenseTime co-founder Lin Dahua: There is a gap with OpenAI, but our costs are only one-tenth of theirs.

SenseTime News: In an interview with CNBC, Lin Dahua, co-founder and Chief Scientist of SenseTime Technology, said that the latest model, Sensenova U1, still has a gap in raw image generation compared with OpenAI’s GPT Image 2 and Google’s Gemini Nano Banana, but the cost of generating images is only one-tenth of ChatGPT Images 2.0. He emphasized that many scenarios do not require the most advanced model—sufficiency is enough. Sensenova U1 is built on the self-developed Neo-Unify architecture, integrating language and visual processing into a single system to improve speed and efficiency. Lin Dahua revealed that ByteDance’s video model Seedance once created competitive pressure, and SenseTime has integrated some of its capabilities into the short-video tool Seko. In terms of finances, SenseTime narrowed its net loss by 58.6% in 2025, and EBITDA turned positive for the first time in the second half of the year. Jefferies noted that pure-model companies face dilemmas such as low customer loyalty and limited differentiation, while platform companies such as Alibaba and Tencent can subsidize AI development using cash flows from core businesses; independent AI companies, however, continue to incur ongoing losses. In response to U.S. sanctions, SenseTime has shifted the focus of its overseas expansion to Southeast Asia, North Asia, the Middle East, and Brazil.