OpenAI recently disclosed that it has no immediate plans to provide an application programming interface (API) for its Sora video generation model. Sora is an AI model capable of creating realistic videos based on textual descriptions or reference images.

In a public Q&A session, Romain Huet, the head of developer experience at OpenAI, clearly stated that there are currently no plans to launch a Sora API. Previously, due to an overwhelming number of users, OpenAI had temporarily closed the application portal for its Sora-based video creation and editing toolkit. In response, OpenAI CEO Sam Altman apologized on social media, acknowledging that they had underestimated the demand for Sora and were working to resolve the access issues. The registration for Sora has since been reopened.

This decision by OpenAI may put it at a competitive disadvantage. Its rival, Google, launched a limited-access API for its video generation model, Veo, in early December and announced plans to release an API for Veo 2 in 2025. Additionally, Amazon Web Services (AWS) has provided an API for its new Nova Reel video model, and several startups specializing in generative video have also made APIs available for their models.

Despite the competitive pressure, OpenAI has chosen not to introduce an API for Sora at this time. This decision may be due to current capacity issues, as OpenAI needs time to address the challenges posed by high user traffic. Whether OpenAI will change its stance and offer a Sora API in the future, as technology advances and demand continues to grow, remains to be seen and anticipated.