OpenAI Expands Sora 2 Video API with Longer Clips, Continuation, and Batch Jobs

OpenAI Developers said on X on March 12, 2026 that the Video API gained several new capabilities powered by Sora 2: custom characters and objects, 16:9 and 9:16 exports, clips up to 20 seconds, video continuation, and batch jobs. That is a meaningful expansion because it turns the API from a basic generation endpoint into something closer to a production workflow tool.

Each of the additions addresses a common bottleneck in video generation. Custom characters and objects help keep assets consistent across multiple shots. Dual aspect ratios matter because teams rarely publish in only one format; marketing, social, and product video pipelines usually need both vertical and horizontal outputs. Extending maximum clip length to 20 seconds and adding continuation also make it easier to build sequences instead of isolated fragments.

The operational angle is just as important. Batch jobs let teams queue larger workloads, which is a better fit for studios, agencies, and product teams generating many variants at once. OpenAI’s Sora 2 model page positions the model as a video system with synced audio, and the official pricing page lists base sora-2 video generation at $0.10 per second for 720x1280 or 1280x720, with higher-priced sora-2-pro tiers above that. That means the new feature set expands creative control, but it also makes cost planning and render orchestration more relevant.

The broader takeaway is that OpenAI is pushing its video tooling toward repeatable development and content pipelines, not one-off demos. When an API supports asset consistency, continuation, multiple aspect ratios, and batch execution, it becomes easier to plug it into real production systems. The X announcement is short, but the product direction is unmistakable: Sora 2 is being packaged as infrastructure for developers, not only as a showcase model.