AI solutions show explosive growth, PATEO (02889) saw its revenue double in the first half of the year, with gross profit increasing by 64.7%.

PATEO (02889) announced its mid-year results for 2026, reporting revenue of approximately 2.228 billion yuan, a year-on-year increase of 105.1%; gross profit of about 257 million yuan, a year-on-year increase of 64.7%; and a net adjusted loss of around 105 million yuan, a year-on-year narrowing of 35.2%. The announcement states that the increase in revenue and gross profit is mainly due to the continuous recognition of the company's core competitive abilities in the market, with a leap forward in the intelligent cockpit solutions segment and explosive growth in AI solutions. In the first half of 2026, AI large models accelerated from technical validation to large-scale production in vehicles, with edge deployment, cockpit-driving integration, and AI agents becoming the focus of industry competition. During the reporting period, the company secured design points for a new generation of intelligent models from a leading OEM in the new energy vehicle sector, leading the provision of the world's first integrated "soft and hard" AI Box solution based on the NVIDIA DRIVE AGX Thor platform, becoming the first "soft and hard" AI application company in the domestic and international markets to reach mass production cooperation with this chip manufacturer. At the same time, the company became one of the first clients signed for Horizon's cockpit-driving integration chip Horizon Starry Sky, co-building a high-cost-performance domestic cockpit integration solution; and collaborated with XUNCE Technology and SAIMO to jointly develop a Token-based physical AI world model and TokenOS, exploring innovative models such as on-board token billing, pushing the transition of on-board AI from hardware sales to a sustainable operation value transfer model. The explosion in demand for AI large models has driven an increase in the demand for high-speed optoelectronic chips and other core components in AI data centers. During the reporting period, the company continued to push forward the acquisition of a Fabless design firm for high-speed optoelectronic chips (whose products are one of the core chips for AI data center optical modules), strategically extending upward into the chip sector and forming a composite layout of "smart automotive solutions + high-performance communication chips," broadening the depth of the "vehicle-cloud" data closed loop and computational power infrastructure. In addition, the company deepened cooperation with NVIDIA around the next-generation computing platform and optical communication, and co-built a "brain-cerebellum" collaborative embodied intelligence system based on open-source HarmonyOS with Shanghai Jiao Tong University, continuously expanding AI capabilities into new scenarios such as Siasun Robot & Automation, and enhancing the AI technology layout across the entire industry chain.