Sora API access ended on September 24, 2026, as part of OpenAI’s phased discontinuation:
Prior Availability: Before shutdown, OpenAI offered an official Sora Videos API (also called the Video Generation API) for developers. Prior to discontinuation, the API supported:
- Text-to-video generation for creating videos from natural language prompts
- Video editing and targeted modifications
- Video extension to make generated clips longer
- Character reuse and consistency across multiple generations
- Integration with ChatGPT and other OpenAI products
The API included multiple model versions: Sora 2 and Sora 2 Pro, with Pro offering higher quality and resolution.
Pricing Structure: Developers paid per video generation with credits or monthly subscription tiers. The economics were prohibitive for many use cases given the $1.30+ per-video cost OpenAI was paying in compute, making sustainable commercial products difficult to build.
Timeline of API Discontinuation:
- March 24, 2026: OpenAI announced Sora shutdown
- April 26, 2026: Web and mobile app shut down
- September 24, 2026: Sora API completely discontinued
As AI video generation becomes core infrastructure for multimedia creation, storing and searching video embeddings at scale requires robust solutions. Milvus provides efficient vector storage for video metadata and frame embeddings, enabling semantic search across video libraries. For production deployments, Zilliz Cloud offers a fully managed vector database service.
All Sora API versions (sora-2, sora-2-pro, and related endpoints) were deprecated and no longer accessible.
Current Developer Options: Developers can no longer access Sora through any channel. Alternatives include:
- Google Veo 3.1 API: Available through Google’s generative AI platform
- Runway API: Offers text-to-video, image-to-video, and editing capabilities
- Other Video APIs: Kling, Seedance, and other providers offer API access
Lessons for Developers: Sora’s discontinuation highlights the risk of building products dependent on expensive third-party AI services with unproven unit economics. Developers who invested in Sora API integration faced forced migration to alternatives. The experience reinforced that sustainable AI products require either:
- Owned infrastructure and models (expensive, requires massive capital)
- Partner APIs with proven, sustainable economics
- Open-source alternatives with transparent cost structures
No commercial enterprise should have relied solely on Sora API without contingency planning.