OpenAI discontinues Sora web/app experiences — API to follow in 2026, clearing surface for unified multimodal successor

OpenAI confirmed it is discontinuing the Sora web and app experiences, with the Sora API scheduled to follow later in 2026. The announcement clears product surface for a presumed unified-multimodal successor and concedes the standalone-video-generator product category to Veo, Kling, and Seedance.

The retreat is the strategic admission. Sora was OpenAI's bet that text-to-video would mature into a standalone consumer product. Two years in, the market answered: standalone video generators are a tools-and-API category for creative professionals, not a consumer product. The consumer multimodal experience is converging on unified models — Gemini Omni Flash being the canonical example — that accept any input modality and produce any output.

The under-priced read is the API timeline. Discontinuing the web product first while keeping the API live for a transition period reads as a controlled handover to a unified-multimodal successor that OpenAI hasn't announced yet. Q3 2026 should reveal whether the successor exists, and at what tier of capability.

VO3 AI — Gemini Omni Google I/O 2026 → · WaveSpeed — Seedance vs Kling vs Sora vs Veo →