What Is the ChatGPT Video Generator Called?
It was called Sora, and OpenAI discontinued it in 2026. Here is what each name people search for actually refers to, what ChatGPT can still do with video, and what to use instead.
By openCanviz • October 27, 2026
6 min read
The name you are looking for is Sora. That was OpenAI's video generator, and it is gone: OpenAI announced on 24 March 2026 that it was discontinuing Sora, switched off the app and website on 26 April 2026, and ended the API on 24 September 2026. There is no OpenAI video generator inside ChatGPT today, and there never was one in the chat window. If you searched this question because an article or a video promised you a "ChatGPT video generator" and you could not find the button, that is why. The button does not exist.
That leaves a more useful question, which is what the name you half remember actually refers to. There are five or six of them in circulation, they mean different things, and two of them are not OpenAI products at all.
The names, and what each one really is
| Name you may have seen | What it actually is | Still available? |
| Sora | OpenAI's text-to-video model and the app built on it | No. App closed 26 April 2026, API closed 24 September 2026 |
| Sora 2 | The second version of that model, launched 30 September 2025 | No, closed with the rest of Sora |
| "ChatGPT video generator" | Not a product name. Usually means a third-party video app connected inside ChatGPT | The apps exist, ChatGPT does not render video itself |
| DALL-E | OpenAI's image generator, built into ChatGPT | Images only, no video, no sound |
| GPT-4o / GPT-5 image tools | The image generation built into recent ChatGPT models | Still images, not video |
| Sites using "Sora" in their name today | Third-party products borrowing the name | Not OpenAI products |
The last row is the one that costs people money. After a well-known model is discontinued its name stays valuable, so unrelated sites keep using it. If a site charges you a subscription for "Sora", you are not buying anything from OpenAI.
So what does ChatGPT actually produce?
On its own, ChatGPT gives you text, still images, spoken replies in voice mode, and files from its code tool. No narrated, edited video comes out of it. What people call a ChatGPT video generator is one of the third-party apps you can connect inside the chat, which renders the video on its own servers using your account and credits with that company. Best ChatGPT video generators, and what they actually do sorts those apps into the four kinds of picture they make, because that is what decides whether the result is useful to you.
If you want the line-by-line version of what ChatGPT can and cannot output by itself, can ChatGPT make a video? is the honest answer.
Why the old articles still say otherwise
Most of the "how to use the ChatGPT video generator" articles were written between early 2025 and early 2026, when OpenAI had said it planned to bring Sora into ChatGPT. That integration was announced and never shipped, and then the product it would have brought in was discontinued. The articles stayed online. So did the videos. The advice in them describes a route that was already closed by the time most people read it.
Dates worth keeping straight, because they are what separates a current article from a stale one:
- 1
30 September 2025
OpenAI launches Sora 2 along with a standalone Sora app, invite only, on iPhone.
- 2
Early 2025 to early 2026
OpenAI says it plans to bring Sora into ChatGPT. Many articles are written as if that already happened.
- 3
24 March 2026
OpenAI announces it is discontinuing Sora in both the app and the API.
- 4
26 April 2026
The Sora app and web experience are switched off.
- 5
24 September 2026
The Sora API is switched off. Nothing OpenAI sells generates video after this date.
What people usually want when they search this
In practice the question splits into two very different jobs, and only one of them was ever served by Sora.
A short generated clip. Something cinematic from a prompt: a drone shot over a coastline, a product turning in light. This is what Sora did, and it is what the generated footage apps still do. The thing to know about this kind of output is that it invents details. Any text, label, number or diagram inside the clip is decoration, not information.
A video that explains something. A lesson, a process, a concept, a training walkthrough. This is what most people are actually after when they ask ChatGPT for a video, and it is the job generated footage is worst at, because the picture has to carry a specific claim rather than a mood. For this, the useful split is to let ChatGPT do what it is genuinely good at, which is drafting and tightening the script, and then make the video in a tool that draws your script rather than illustrating its general vibe.
That second route is what openCanviz does: you paste the script or the document, it drafts a narrated video scene by scene, and you edit any scene that does not show what its sentence says. The walkthrough, including what to strip out of a ChatGPT script before you paste it, is in how to turn a ChatGPT script into an animated video.
Common questions
What is the ChatGPT video generator called? Sora. It was OpenAI's video model and app, and OpenAI discontinued it in 2026. ChatGPT itself has no video generator.
Is Sora coming back? OpenAI has not announced that it is. As of October 2026 both the app and the API are closed, and no ChatGPT plan includes a video model.
Is there a free ChatGPT video generator? Not from OpenAI. The third-party apps you can connect in ChatGPT have their own free tiers and their own credits, charged by that company rather than by OpenAI. Check which account the rendering is billed to before you plan a project around one.
What should I use instead for a lesson or a training video? Write and check the script in ChatGPT, then make the video in a tool that builds scenes from that script. For an explanation the picture has to show the thing you are claiming, which is the one thing a generated clip cannot promise.
Does ChatGPT have Sora built in now? No. The integration was announced in early 2025 and never shipped, and the product was discontinued in 2026.
You already have the script
Paste the script ChatGPT wrote into openCanviz and it drafts a narrated video, one scene per point, that you can edit scene by scene. It is free to start.
Turn any concept into an animated explainer
Type an outline, get a narrated, animated whiteboard video in minutes. No design skills, no timeline scrubbing. Free to start.
Keep reading
A language lesson video teaches one grammar point or one small set of words, in context, with every example said twice and shown in writing. How to pace it, what to put on screen, how to handle two languages, and a full scene list for the Spanish preterite and imperfect.
A case study video runs situation, problem, options, recommendation, with evidence in every scene and a decision at the end. How to structure it, compare options fairly, keep hindsight out of the analysis, and a worked script on Netflix splitting off its DVD business as Qwikster in 2011.
An equation in a video should arrive one symbol at a time, each with its meaning and its unit, before any numbers go in. How to reveal it term by term, read it aloud, check the units and finish with a worked number, using the ideal gas law PV = nRT.
Untuk siswa SMP dan SMA: cara mengubah tugas presentasi atau proyek menjadi video penjelasan berbahasa Indonesia yang dinilai dari pemahaman, bukan efek. Durasi per jenjang, contoh lengkap tentang gunung api di Indonesia, soal suara, dan aturan AI.