Gasgoo Munich- On July 31, Volcengine officially unveiled Seedance 2.5, its next-generation AI video generation model. To streamline integration for businesses, the company plans to open the model's enterprise API access in the near future.
Seven companies—including XPENG Motors, AgiBot, XCMG Group, PsiBot, Differential Robotics (Hangzhou) Technology Co., Ltd., Noematrix, and Xspark AI—have confirmed partnerships, becoming the first batch of domestic clients to deploy the model.
Image source: Volcengine
Developed in-house by ByteDance and commercialized through Volcengine, the Seedance series sees its latest version extend single-generation video length from 15 to 30 seconds. It also enhances multi-material referencing and precise editing capabilities. This balances creative completeness with control, catering to the demand for high-volume commercial video production across various industries.
While AI video tools were previously used mostly for short-form entertainment and advertising, this batch of partners spans automotive, construction machinery, robotics, and intelligent aviation. The shift signals that video generation AI is breaking out of the media and entertainment bubble, penetrating into industrial R&D and corporate operations.
Clients across sectors have distinct deployment plans. XPENG Motors, for instance, plans to integrate Seedance 2.5 into its internal design system. The automaker will use it to rapidly generate vehicle visualization assets and multilingual explainer videos, cutting production timelines for new model launches and overseas marketing materials.
Embodied intelligence firms like AgiBot and Noematrix can use the AI to mass-generate robot motion scenarios, supplementing the video data needed for algorithm training without relying on repeated physical filming. Meanwhile, manufacturers like XCMG are targeting industrial training, using the tech to produce equipment operation tutorials and lower the costs of live-action shooting and 3D animation.
For enterprises, the open API model allows AI video capabilities to be embedded directly into internal office, R&D, and design systems. This eliminates the need to switch platforms, suiting the demand for high-volume, high-frequency video production. Covering mobility, heavy industry, and robotics, these pioneering partners demonstrate that the application scenarios for general-purpose video models are widening—moving well beyond new media content creation.
