OpenAI's Sora 2 video generator lasted less than seven months before the company shut down the app, website, and scheduled the API for closure in September 2026. The model, which launched September 30, 2025, with native synchronized audio and physics-aware realism, became a casualty of economics rather than technical failure. The discontinuation reveals how even cutting-edge AI products can struggle when compute costs spiral and user engagement collapses.

What Made Sora 2 Different From Other Video Models?

Sora 2 introduced two capabilities that reset expectations for the entire video generation category. First, it modeled physics honestly. When a basketball shot missed, it would rebound off the backboard rather than teleporting into the hoop. The model could handle complex dynamics like backflips on a paddleboard that accurately reflected buoyancy and rigidity. Earlier video models tended to warp reality to satisfy the prompt; Sora 2 let physical outcomes play out naturally.

Second, Sora 2 generated dialogue, sound effects, and ambient sound together with the video in a single pass, rather than requiring separate audio synthesis. This was the headline change from every prior version of Sora and made single-generation clips feel like finished footage instead of silent b-roll. The model also gained genuine creative champions. Filmmaker Paul Trillo, who directed a commissioned Sora music video, said the experience felt liberating.

Why Did OpenAI Shut Down Sora 2?

Economics was the primary driver. Reporting sourced to the Wall Street Journal estimated Sora's compute costs at roughly one million dollars per day while usage collapsed from around one million users at peak to under 500,000. The app fell from number one on the App Store to number 165 by March 2026. OpenAI declared the free tier "completely unsustainable" barely a month after launch and began charging for extra generations in November 2025, just 32 days after the September 2025 debut.

The strategic backdrop was a pivot toward coding, enterprise, and agent products. The Disney partnership, announced with a one-year exclusivity in December 2025 when Disney invested one billion dollars, ended roughly fifteen weeks later when the shutdown was announced. Notably, OpenAI never published a blog post about the discontinuation; the news went out via an X post and a help-center FAQ.

What Were Sora 2's Actual Specifications?

Sora 2's specs are among the most misquoted on the internet, largely because the app and the API had different limits, and the API quietly gained capabilities that were only documented in OpenAI's developer cookbook. The consumer version supported up to 1080p resolution and 25 seconds of continuous video for Pro users on the web via storyboards. The API version supported higher resolutions, including 1024x1792, 1792x1024, and 1080x1920 formats.

A widespread claim that "Sora 2 maxes out at 25 seconds" is incorrect for the API. OpenAI's Sora 2 prompting guide documented 16- and 20-second generations plus extension chaining to a true 120-second maximum, a capability that outlived the consumer app by five months. The API also shipped a Characters API that built reusable characters from 2 to 4 second reference clips, with human likenesses gated behind account-manager approval.

How to Transition From Sora 2 to Alternative Video Models

• Evaluate Current Alternatives: Veo 3.1, Kling 3, Seedance 2.0, and MiniMax-H3 now cover the jobs Sora 2 used to do, each with different strengths in physics simulation, audio synchronization, and resolution capabilities.
• Understand API Pricing Differences: Sora 2 API pricing ranged from $0.10 per second for the base model to $0.70 per second at 1080p for the Pro version; compare these rates against competing models before migrating workflows.
• Export Your Existing Content: Videos already exported from Sora 2 remain yours to keep and use; the ability to generate new ones through OpenAI ends September 24, 2026, so plan your content pipeline accordingly.
• Leverage OpenAI's Prompting Guide: OpenAI's Sora 2 prompting guide remains one of the best short documents on directing a video model, and its advice transfers directly to current models, so study its cinematography principles before switching platforms.

What Happened to Sora's Timeline From Launch to Shutdown?

The complete arc reveals a product in distress from early on. Sora was unveiled as a research preview in February 2024, generating up to 60-second clips framed as "video generation models as world simulators," but was never publicly shipped in that form. Sora Turbo launched at sora.com in December 2024 as a ChatGPT Plus and Pro perk, supporting up to 1080p and 20 seconds with tools like Storyboard, Remix, Blend, and Loop, but without audio.

Sora 2 and Sora 2 Pro launched September 30, 2025, alongside an invite-only iOS app in the US and Canada with a short-video feed and Cameos. OpenAI called it a possible "GPT-3.5 moment for video." The API arrived October 6, 2025, at DevDay. By November 2025, the Android app expanded to Japan, South Korea, Taiwan, Thailand, and Vietnam, but OpenAI introduced paid extra generations just 32 days after launch, calling the free tier "completely unsustainable." Disney's one-billion-dollar investment and character licensing came December 11, 2025. Sora 1 was sunset for US users March 13, 2026, with permanent library deletion. The discontinuation announcement came March 24, 2026, via X post and FAQ, with Disney exiting the partnership. The app and sora.com shut down April 26, 2026, and the API is scheduled to end September 24, 2026.

Did Sora 2 Videos Have Watermarks?

Yes, but there were two different regimes, and most explainers conflate them. Consumer videos downloaded from the app and sora.com carried a visible moving watermark plus C2PA Content Credentials metadata. OpenAI also ran internal reverse video and audio search to trace outputs. API outputs carried C2PA metadata only, with no visible watermark. Watermark-removal tools circulated within a week of the Sora 2 launch, which is why C2PA metadata, not the visible mark, became the more durable provenance layer.

The Sora 2 shutdown marks a turning point in how the AI industry approaches consumer-facing generative products. While the technology itself proved capable of producing impressive results, the economics of running a free or low-cost video generation service at scale proved untenable. For creators who relied on Sora 2, the transition to alternative models requires evaluating not just feature parity but also pricing, API stability, and long-term company commitment to the product.