OpenAI says it won't be releasing an API for its video-generating model, Sora, anytime soon. OpenAI's Sora model launched during its holiday-themed 12-day "shipmas" event and it got a lot of attention.

In case you're not familiar, Sora is a powerful text-to-video model developed by OpenAI. The model allows users to create videos by just providing text or images which is a big jump in AI video generation. It can generate videos in 1080p resolution, up to 20 seconds long, and in widescreen, vertical, or square formats.

However, despite the popularity and demand, OpenAI isn't ready to release an API yet. During an AMA session, Romain Huet, head of developer experience at OpenAI, said "We don't have plans for a Sora API yet, but we'd love to hear more! What will you build if we ship one?". It seems the company is working on other things and will consider future developments based on user feedback.

OpenAI probably isn't focusing on the API right now because of capacity issues. It seems the demand for Sora is so high that OpenAI can't manage the influx of users. After Sora was released OpenAI had to shut down new signups for its video generation tools because the traffic was way higher than expected. OpenAI's CEO, Sam Altman, apologized for the inconvenience on X.

OpenAI is behind some of its competitors in the AI video generation space. Google for example has already released an API (limited access) for its video model, Veo (Veo can create AI-generated videos). Google has announced an API for the new version, Veo 2, sometime in 2025.

Amazon Web Services (AWS) has released an API for its latest video model, Nova Reel. Several other generative video startups including Runway also have APIs. Runway says its API is used by some of the biggest consumer tech companies to create millions of videos for their users.

OpenAI will likely need to reassess its approach if it wants to stay competitive, especially with other companies like AWS and startups offering APIs for similar video generation tools.