How to use Sora?
You can't use Sora anymore: OpenAI discontinued the app and website on April 26, 2026. When live, you signed in with an OpenAI account, typed a text prompt (or uploaded an image), picked settings, and Sora generated a short video with audio. Today, use Google Veo, Runway, Kling, or Pika instead.
Why — the first-principles explanation
Using Sora was designed to feel as simple as typing a text message, and that was the whole point of a prompt-based interface: you describe what you want in plain language, and the model handles the hard work of turning words into moving pictures. Behind the scenes it converted your prompt into a video by denoising random noise into frames that matched your description, but you never saw that — you just saw a text box and a 'generate' button.
In practice, using Sora 2 meant opening the app or sora.com, signing in, and writing a clear prompt like 'a red kite over a beach at sunset, camera slowly rising.' You could also upload a starting image to animate, use the Cameo feature to place a consented likeness into a scene, remix others' clips from the feed, and adjust length or aspect ratio. Better prompts — with subject, action, setting, and camera movement — produced better results.
The important reality in 2026: this is now historical. OpenAI shut down the Sora app and website on April 26, 2026, so there is no interface left to use. The only remnant is the developer API (ending September 24, 2026), which requires coding rather than a friendly app. For everyday video creation, the practical move is to learn a currently active tool.
An example that makes it click
Using Sora was like ordering at a sandwich counter that builds anything you describe. You'd say, 'a turkey club on rye, toasted, cut in triangles,' and a finished sandwich slid out. With Sora you'd say, 'a puppy chasing bubbles in slow motion,' and a 10-second video slid out. But the counter closed in April 2026 — so now you have to order from a different shop, like Veo or Runway.
How to do it
- Note the status first: as of 2026 the Sora app and sora.com are shut down, so these steps describe how it worked and no longer apply.
- When live: open the Sora app or sora.com and sign in with your OpenAI account.
- Type a detailed prompt including subject, action, setting, and camera movement — or upload a starting image.
- Choose options like clip length and aspect ratio, then tap Generate and wait for the render.
- Review, remix, or extend the clip, then download it (outputs carried a visible watermark).
- Today, use an active tool such as Google Veo, Runway, Kling, or Pika.
Key facts
- Sora used a plain-language prompt box; no editing skills were required.
- It accepted text-to-video and image-to-video inputs.
- The Cameo feature let users insert a consented likeness into scenes.
- All outputs carried a visible watermark and C2PA provenance metadata.
- The app and website were discontinued April 26, 2026.
OpenAI's text-to-video model for short cinematic clips.
Affiliate link — we may earn a commission at no cost to you.▶ The 60-second explainer (script)
How do you use Sora? Here's the thing, you can't anymore. OpenAI shut down the Sora app and website on April 26, 2026. But when it worked, it was simple. You opened the app, signed in with your OpenAI account, and typed a prompt describing your video, like 'a red kite over a beach at sunset.' You could also upload a photo to animate, or use the Cameo feature to drop a consented face into a scene. Then you picked a length, hit generate, and a short video with sound appeared, complete with a watermark. The clearer your prompt, the better the result. Since Sora is gone for regular users, today you'd reach for a tool like Google Veo, Runway, Kling, or Pika to do the same kind of thing.
What authoritative sources say
People also ask
Is there a Sora app to download?
No longer. The iOS and Android apps were removed when Sora shut down on April 26, 2026.
Did you need coding skills to use Sora?
No. The app used a plain-language prompt box; only the developer API required code.
Could Sora animate a photo?
Yes, it supported image-to-video: you uploaded a starting frame and it generated motion from it.
What can I use instead?
Active AI video tools in 2026 include Google Veo, Runway, Kling, and Pika.