OpenAI added five capabilities to its Video API, powered by sora-2: custom characters and objects for visual consistency across generations, 16:9 and 9:16 aspect ratio exports, clips up to 20 seconds, video continuation to extend completed scenes, and batch jobs for large offline render queues. The API supports two models — sora-2 for speed and rapid iteration, and sora-2-pro for production-quality 1080p output.
This brings programmatic control over Sora 2's full feature set to all developers. Studios, brands, and content teams can now build video generation pipelines at scale — from campaign creative to storyboarding and UGC production.
Submit a large offline render queue through the Batch API for studio workflows, or reuse character assets across multiple generations to maintain visual consistency throughout a project.