OpenAI currently has no plans to release an API for its AI video-generation model, Sora. The model, which can create realistic videos from text descriptions or images, is facing capacity challenges.
During a recent AMA with OpenAI’s development team, Romain Huet, head of developer experience, made it clear: “We don’t have plans for a Sora API yet.”
The issue appears to be high demand. OpenAI had to pause applications for its Sora-powered video tools shortly after launch due to unexpected traffic spikes. CEO Sam Altman addressed the situation on X, saying, “We significantly underestimated demand for Sora. It’s going to take a while to get everyone access. Trying to figure out how to do it as fast as possible!”
Sign-ups for Sora resumed just a few days ago. However, OpenAI’s decision not to prioritize an API may put it at a disadvantage. Competitors like Google and AWS are moving ahead with APIs for their video models. Google’s Veo launched an API in December, and its upcoming Veo 2 will follow in 2025. AWS has already released an API for its Nova Reel model.
Startups like Runway are also stepping up. Runway claims its API has powered “millions of videos” for major tech companies.
For now, OpenAI’s focus seems to be on managing Sora’s overwhelming demand, not expanding its access through an API.
Published: December 18, 2024