Sora 2 was OpenAI's flagship video and audio generation model, released on September 30, 2025, alongside a standalone social app called Sora. It represented one of the biggest leaps in AI video generation at the time — more accurate physics, synchronized audio, and a feature that let people insert themselves into AI-generated scenes. This guide covers what Sora 2 was, what it could do, and its current status, since the product's availability has changed significantly since launch.
Important Update: Sora's Current Status
Before going further, here's the most important thing to know: the Sora consumer app and web experience were discontinued on April 26, 2026. OpenAI cited unsustainable operating costs relative to revenue as the primary reason, and the company redirected its focus toward coding and enterprise products. The Sora API, used by developers to access the sora-2 and sora-2-pro models directly, was discontinued on September 24, 2026.
That means as of today, the consumer Sora app and sora.com are no longer accessible, and the underlying API was shut down on September 24, 2026. If you're researching Sora 2 to actually create AI video today, it's worth knowing this upfront rather than discovering it after signing up.
What Was Sora 2?
Sora 2 was OpenAI's second-generation video model, following the original Sora released in February 2024. It was designed to generate video from text prompts, images, or existing video clips, with a stronger emphasis on realistic physics and world simulation than prior models.
The model powered two things:
The sora-2 and sora-2-pro API models, used by developers and third-party platforms
A standalone Sora iOS app, a social, TikTok-style feed built specifically around Sora-generated content
Key Features Sora 2 Introduced
Improved Physics and Realism
Sora 2 was built to better follow real-world physical behavior. OpenAI highlighted examples like a missed basketball shot rebounding off the backboard instead of teleporting into the hoop, and paddleboard scenes that modeled water and buoyancy realistically.
Synchronized Audio
Unlike many earlier video models that produced silent clips, Sora 2 generated audio alongside video — background sound, dialogue, and effects synced to the visuals in the same generation pass.
Cameos and Characters
One of Sora 2's most talked-about features let users record a short video and audio sample to create a verified likeness, which could then be inserted into any Sora-generated scene. OpenAI built in consent controls so people could decide who was allowed to use their likeness and revoke access at any time.
Multi-Shot Consistency
Sora 2 could follow instructions across multiple shots while maintaining consistent characters, settings, and world state, rather than treating each shot as an isolated generation.
The Sora App
The standalone Sora app packaged these capabilities into a social, feed-based experience, letting users create, remix others' videos, and discover content through an algorithmic feed, similar in structure to TikTok or Instagram Reels.
Why Sora Was Discontinued
According to reporting following the shutdown announcement, Sora's operating costs were substantial relative to the revenue it generated, with estimates citing daily inference costs in the range of $1–15 million against a total lifetime revenue of roughly $2.1 million from in-app purchases. Active users reportedly declined from an early peak toward the low hundreds of thousands in the months following launch. A previously announced Disney content partnership, which would have licensed over 200 characters for use in Sora, was also abandoned around the same time as the shutdown decision.
OpenAI has said Sora will continue to exist as an internal research effort focused on world models, but has not announced a public successor product as of this writing.
What Happens to Content Made With Sora?
OpenAI directed users to export any content stored in their Sora library before the April 26, 2026 cutoff. Data associated with Sora accounts that wasn't exported in time is subject to permanent deletion once the discontinuation process completes. If you have Sora-generated content you still need, checking OpenAI's Help Center for the most current export guidance is the safest next step.
What to Use Instead of Sora 2
Since Sora is no longer available as a consumer product, creators looking for text-to-video generation now need a different platform. Rather than relying on a single-vendor tool, many creators are moving toward platforms that offer access to multiple AI video models in one place, which reduces the risk of being caught off guard by another single-product shutdown.
TextVidai works this way you can generate AI videos from text and choose from multiple AI video models rather than depending on one provider. If you want to see what's possible before creating your own, the AI-generated video examples on the showcase page show a range of styles across different models. You can also review TextVidai pricing to compare plans.
Frequently Asked Questions About Sora 2
What is Sora 2?
Sora 2 was OpenAI's flagship video and audio generation model, released September 30, 2025, known for improved physics accuracy, synchronized audio, and a "cameo" feature that let users insert their likeness into generated scenes.
Is Sora 2 still available?
The Sora consumer app and web experience were discontinued on April 26, 2026. The Sora API was then discontinued on September 24, 2026, so the sora-2 and sora-2-pro models are no longer accessible through any channel.
Why did OpenAI shut down Sora?
Reporting points to unsustainable operating costs relative to revenue, a strategic shift toward coding and enterprise products ahead of a potential IPO, and the collapse of a previously announced Disney content partnership.
What features did Sora 2 have?
Sora 2 introduced improved real-world physics, synchronized audio and dialogue generation, multi-shot consistency, and a cameo feature for inserting verified likenesses into AI-generated scenes.
Can I still access videos I made with Sora?
OpenAI advised users to export their content before the April 26, 2026 shutdown. Content not exported in time is subject to permanent deletion once the discontinuation process is complete; check OpenAI's Help Center for the latest guidance.
What can I use instead of Sora 2 now?
Platforms that support multiple AI video models, such as TextVidai, offer an alternative that isn't dependent on a single provider's product roadmap.
Conclusion
Sora 2 was a genuine step forward for AI video when it launched in September 2025, introducing synchronized audio, improved physics, and the cameo feature that made it briefly one of the most talked-about consumer AI products. But as of April 26, 2026, the Sora app and web experience are no longer available, with the API following in September 2026. For creators who relied on Sora, or anyone researching it now, the practical next step is choosing a platform that isn't tied to a single model's future you can use TextVidai to generate AI video across multiple models instead.