Sora lands inside ChatGPT
OpenAI is reportedly folding its Sora video-generation tech directly into the ChatGPT chat interface, letting users create short videos from text prompts inside the same app reported. That will speed ideation and rough-cut prototyping for editors and producers — but also risks bigger compute costs and wider non‑specialist use that could compress demand for high-end finishing work.
The Information reported theinformation.com on March 11, 2026 that OpenAI is planning to fold its Sora text‑to‑video model directly into the ChatGPT app. OpenAI’s developer docs list Sora 2 with a $0.10/second rate for the standard 720p model and show portrait (720×1280) and landscape (1280×720) outputs with synced audio as supported outputs. developers.openai.com OpenAI’s Sora product pages state the Sora app can publish videos up to 1080p and up to 20 seconds long, with options to extend or remix uploaded footage. openai.com Simple math from the $0.10/second standard rate implies a 10‑second 720p draft costs roughly $1, while third‑party reporting of Sora 2 Pro pricing at about $0.30–$0.50/second yields a 10‑second HD clip costing roughly $3–$5. developers.openai.com OpenAI announced ChatGPT reached more than 900 million weekly active users on February 27, 2026, a scale that would expose any embedded video generator to a vastly larger non‑specialist audience. openai.com Engadget documented that the standalone Sora app slipped out of the App Store top‑100 and noted OpenAI had offered roughly 30 free Sora generations per account per day in 2025, signaling both adoption limits and existing free‑tier credits. engadget.com Reporting summarized from The Information tied the integration to OpenAI’s internal modeling that inference costs could top $225 billion between 2026 and 2030, a figure that would materially change forecasting for enterprise post‑production compute budgets. theinformation.com Microsoft’s Foundry documentation and OpenAI notes show Sora 2 supports “remix” workflows that make targeted edits to existing clips rather than full regenerations, a capability that can shift labor from traditional offline conform and manual VFX to iterative prompt‑driven refinements. learn.microsoft.com
