OpenAI is reportedly closing its standalone Sora video generation app, marking a significant pivot in its product strategy.
This move appears to be a calculated business decision rather than a failure of the technology itself. The core reasons can be broken down into three main areas.
First, the unit economics were challenging. Industry analysis estimated that generating a single 10-second video clip with Sora cost around $1.30 in computing power. However, user spending was very low, with a lifetime average of only about $0.15 per download. With app installs and revenue declining sharply in early 2026, sustaining a costly consumer app became difficult.
Second, OpenAI is shifting its product strategy. Reports from early March indicated a plan to integrate Sora's capabilities directly into ChatGPT. This consolidates services into a "super-app," making the standalone Sora app redundant. This move allows OpenAI to offer video generation to higher-paying ChatGPT Pro subscribers and enterprise clients, where monetization is much stronger.
Finally, mounting legal and regulatory risks played a significant role. The rise of deepfakes, pressure from rights holders over copyrighted training data, and upcoming regulations like the EU AI Act created a complex and costly environment for a mass-market video app. By moving video generation into more controlled environments like ChatGPT and enterprise APIs, OpenAI can better manage compliance and safety.
In essence, shutting down the Sora app isn't an admission of defeat. It's a strategic retreat from the volatile consumer social space toward a more sustainable, profitable, and controllable B2B and premium-subscriber model.
- Glossary -
- Unit Economics: The direct revenues and costs associated with a particular business model on a per-unit basis. For Sora, this means the cost to generate a video versus the revenue earned from a user.
- Inference Cost: The computational cost incurred when an AI model runs to generate an output (like a video) based on a user's prompt.
- ARPU (Average Revenue Per User): A metric that measures the average revenue generated from each active user over a specific period.