The short answer
No, not by itself. As of November 2026, ChatGPT writes text, reads files, talks and makes images. It doesn't render video files. OpenAI's video model, Sora, was the closest thing, and OpenAI shut it down this year.
That's not the whole story, though. ChatGPT is good at the parts of a video that take the most thinking: the script, the order of scenes, what goes on screen and what gets said. And plugins can now turn a conversation into something you watch without leaving the chat. Which route fits depends on the kind of video you want.
What happened to Sora
OpenAI announced on 24 March 2026 that it was discontinuing Sora. The Sora app and website closed on 26 April 2026, and the Sora API followed on 24 September 2026. OpenAI's help center has a page called "What to know about the Sora discontinuation" if you had videos stored there.
So any guide that tells you to "just ask ChatGPT for a video" and points at Sora is out of date. There's no built-in video button in ChatGPT today. If OpenAI ships a new video model, this page will change.
One thing that confuses people is that ChatGPT's voice mode can use your phone's camera on some plans. That's video going in, so ChatGPT can see what you point at. Nothing comes out as a video file.
What ChatGPT can do for a video today
Even without rendering, ChatGPT can do most of the planning. Ask it for:
- A script at a set length. Spoken English runs at about 130 to 160 words a minute, so "a 60-second script" means roughly 150 words. Say who it's for.
- A scene list where each scene has one idea, what's on screen, and the line that's said over it.
- On-screen text short enough to read in two seconds.
- Captions split into lines that match the script.
- Still images for a thumbnail or a key frame, made with ChatGPT's image generation.
- Prompts for other tools if you plan to use a separate AI video generator.
The scene list is the most useful of these, and the easiest to skip. A good one stops you from making a video that's just a talking head reading a wall of text.
Three ways to get an actual video out of a chat
1. Copy the script into a video tool
Ask ChatGPT for the script and scene list, then paste them into a video editor or an AI video generator. This works for anything, including ads, social clips and talking-avatar videos. The cost is switching apps, re-explaining the context, and usually paying for another subscription.
2. Use a video plugin inside ChatGPT
ChatGPT supports plugins, which OpenAI called "apps" until mid-2026. Several video companies offer one that takes your script or footage from the chat and sends it to their editor. These suit marketing and social video. Most need an account and credits with the video company.
3. For learning, a narrated lesson that plays in the chat
If the video you want is an explanation of a concept, a process or something you've been discussing, you need a different tool. You want short scenes, a voice, captions, and something that checks you followed along.
That's what Lesson Video does. You ask ChatGPT for a lesson about what you've been talking about. ChatGPT writes the scenes and the narration, and the lesson plays inside the chat with a voice, captions and animation. At the end there's a question, and your answer goes back to ChatGPT so it can pick up where you got stuck.
It's not cinematic footage. Scenes are text, diagrams and simple animation, closer to a good whiteboard explainer than a film. For learning, that's usually what you want anyway.
How to make a lesson video from a conversation
Here's the flow with Lesson Video once it's added to ChatGPT. The setup steps are on the home page.
- Talk it through first. Ask your questions in ChatGPT as you normally would, whether it's a physics idea, a function you don't understand or a historical event. The conversation is what the lesson is built from, so a few back-and-forths help.
-
Ask for the lesson. Something like:
Make this a short lesson I can watch.
Or be specific:
Make a 2-minute lesson on why ice floats, for someone who knows what density is but not hydrogen bonds.
- Watch it in the chat. Captions start right away and the voice follows.
- Answer the question at the end. If you get it wrong, ask ChatGPT to explain that part differently.
- Ask for changes. "Use a 50 cm³ cube in the example instead" changes only the scenes that need it.
The free plan includes 3 lessons a month. Pro is €9 a month for 60.
Prompts that get better video scripts
These work whether you render in ChatGPT with a plugin or paste into another tool.
- "Write this as 6 scenes. Each scene gets one idea, at most 12 words of on-screen text, and one or two spoken sentences."
- "Start with a question or a surprising fact, not a definition."
- "Use one concrete example and stick with it for the whole video."
- "End with one question that checks whether the viewer understood the main point. Don't make it a trick question."
- "Read the script back and cut every sentence that repeats an earlier one."
The last one matters more than it looks. First drafts of narration repeat themselves a lot, and repetition feels much slower when it's spoken than when you read it.
Which option should you pick?
- Ads, social clips, product demos. Write the script in ChatGPT, then use a video tool or a video plugin.
- Talking-avatar or presenter videos. Use a dedicated avatar tool. ChatGPT writes the script.
- Explaining something to yourself or a class. Use a lesson plugin like Lesson Video, so the explanation, the narration and the check question all come from the same conversation.
- A one-off image or thumbnail. ChatGPT's own image generation is enough.