Skip to main content

Can ChatGPT Generate Video? What Still Works in 2026

·
9 min read
·
By Leila Morgan
·
AI in Film
A linocut-style illustration of a dark desk under a lamp, with faintly marked script pages and a mug on the left and three small storyboard frames on the right: a beekeeper checking a wooden hive, looking up in a storm while holding a hive box, and carrying a box down a stairwell

Can ChatGPT make videos? What it can and cannot do

People searching “can ChatGPT generate video,” or the close question “can ChatGPT make videos,” want a clear yes or no. As of November 2026, the answer is no: ChatGPT cannot generate a video file today. It is good at scripts, shot lists, prompts, captions and voiceover text, and it can still make still images. Looping GIF creation is rolling out in ChatGPT, but a GIF is not a substitute for a shot. If a clip appears inside the chat, it almost always comes from a third-party app, not from ChatGPT itself. A filmmaking studio is how you turn that plan into storyboarded shots.

ChatGPT is not a dedicated video generator, and OpenAI no longer offers a built-in video model. That was not always true: Sora, OpenAI’s video product, used to generate clips. The Sora web and app experiences were discontinued on April 26, 2026, and the Videos API and Sora 2 models were shut down on September 24, 2026, with no one-to-one replacement. The dates, and what ChatGPT can still produce, are below.

TaskWhat ChatGPT can do today
Scripts, shot lists, prompts, captions and voiceover textYes
Still imagesYes. ChatGPT Images (Images 2.6) is available on every plan
Looping GIFsRolling out. A GIF is not a film shot or a video file
A video file, such as an MP4No
A built-in OpenAI video modelNo. Sora’s consumer product and API are shut down
Editing a video fileNo native editor. It can help you plan cuts
A clip that appears inside the chatOnly through a third-party app, GPT or plugin, on that company’s model and plan
A linocut-style work table under a desk lamp, with faintly marked script pages, a portrait of a beekeeper and a fan of stills of traditional skep hives on the left, and an empty film reel and a blank picture frame on the right
What ChatGPT can still make: a script, a still and, as the feature rolls out, a looping GIF. What it cannot: a video file or a finished sequence of shots. Fictional scene.

ChatGPT Images can already make stills on Free, Go, Plus and Pro, and Images with thinking is available on Plus, Pro and Business (OpenAI says it is coming to Enterprise and Edu). That is useful for concept art, frames and references. It is not the same as generating video. OpenAI’s Images in ChatGPT help article is also where it says GIF creation is rolling out: you can ask for a looping GIF, or upload a still and ask ChatGPT to animate it. ChatGPT cannot edit an animated GIF you upload. Until the feature reaches an account, the honest line is “rolling out,” not “everyone can do this.”

What happened to Sora, OpenAI’s video tool

Sora was OpenAI’s video product. For a time, people could generate clips in the Sora web and app experiences, and developers could call the Videos API with the Sora 2 models. That path is closed. OpenAI has not published a reason in its help article, and this post does not invent one.

  • April 26, 2026: the Sora web and app experiences were discontinued. OpenAI’s Sora discontinuation help article states that date. The old Sora index page carries a banner that the product is no longer available as of April 26, 2026.
  • September 24, 2026: the Sora 2 models and the Videos API were shut down and are no longer available. OpenAI’s video generation API guide says there is no one-to-one replacement API. The deprecations list covers the Videos API, sora-2, sora-2-pro and their dated snapshots, with a notice date of March 24, 2026.
  • Historical only: the last models were sora-2 and sora-2-pro. The retained guide says both supported 16- and 20-second generations, and that sora-2-pro was the 1080p option (1920×1080 or 1080×1920). Those figures are not current access, plan limits or regional availability, because the product and API are shut down.
  • Leftovers: OpenAI says you can export Sora content at sora.chatgpt.com/sunset, that data will be permanently deleted after any final export window, and that purchased ChatGPT or Sora credits can still be used for Codex.

No current ChatGPT plan lists video generation as a feature. The comparison on ChatGPT’s pricing page lists image generation, including Images with thinking, and does not include a video-generation row. Older Plus or Pro allowances from Sora’s launch posts no longer apply.

Can you make a video inside ChatGPT with an app or plugin?

Yes, in a limited way, and the distinction matters. ChatGPT can run third-party apps, GPTs and plugins. Some of those tools will generate or assemble a clip from inside the chat. The file is produced by that company’s model and billed on that company’s plan.

That is not the same as ChatGPT generating the video. It is closer to opening another editor while you are still in the thread. The usual result is a one-off social clip, a stock-footage edit or a short promo, not a planned sequence of shots. If a tutorial says “make a ChatGPT video,” check whether the renderer is ChatGPT or a plugin.

There is also no official ChatGPT video generator to install. Searches for a built-in tool still surface old Sora pages and plugin listings. The current fact is simpler: ChatGPT will write the plan and make stills; the video comes from somewhere else.

Can ChatGPT edit video?

No. ChatGPT has no native video editor. It cannot open an MP4, mark an in-point, lay a cut, or export a revised file. What it can do is help you plan the edit: a shot list, a suggested order, notes on what to trim, and caption or voiceover text to take into an editor.

Some third-party apps claim you can edit from inside the chat. If that works, you are editing on that app’s own timeline, with that app’s credits. It is still not ChatGPT cutting the video. For a finished piece, you still leave the chat and finish in an editor.

Can ChatGPT turn a photo into a video?

Not into a film shot. ChatGPT can generate or edit a still, and, where GIF creation has rolled out, you can upload a photo and ask it to animate a loop. That moving image is a GIF, not an image-to-video clip from a video model, and it is not a substitute for a shot in a sequence.

Image-to-video belongs in a video model: you start from a frame and generate a clip with camera move, motion and duration. If you are writing prompts for that step, see our guide to AI video prompts. The same limit applies to text-to-video: ChatGPT can write the prompt, but it cannot render the clip.

From a ChatGPT plan to a storyboard and generated shots

Use ChatGPT for the plan, then get the plan out of the chat. For script craft, see whether ChatGPT can write a movie script. The full filmmaking workflow, from idea to shots you can edit, is in our guide on how to use ChatGPT for videos. This section is only the handoff.

In CinemaDrop you can paste a premise into Script Wizard or paste a script into a new storyboard. Script Wizard’s Concept step returns five concept directions (15 credits), Outline builds a beat outline (10) and Script writes the screenplay (20). That is 45 credits, inside the free Lite plan’s 50 one-time credits. The Elements step then extracts characters, locations and key props (55 credits, more than Lite has left after that run). You @-tag those Elements in shot prompts so each frame can reuse the same references.

Three linocut-style storyboard frames, two on top and one below, of a fictional rooftop beekeeper: checking wooden hive boxes under dark clouds, looking up as hail begins, and carrying a hive box down a stairwell with a neighbor
A plan taken out of the chat and onto a board. Top: the beekeeper checks the hives under a darkening sky, then looks up as hail begins. Below: she and a neighbor carry a hive box down a stairwell. Fictional scene.

Generate each shot on a model CinemaDrop actually sells. Kling 4.0 is available on Basic and higher (3–30 seconds, 720p or 1080p). Seedance 2.5 needs Plus or higher (4–30 seconds, 1080p, 720p or 480p). Lite can build the script, the storyboard and stills, but it cannot generate video. Basic is $8 a month for 700 credits; Plus is $19 a month for 1,800 credits. Export is a ZIP of storyboard assets, on Basic and up. There is no timeline in the app and no compiled MP4, so you finish in an editor such as CapCut, Premiere or DaVinci. For a longer path from first idea to a short, see how to make an AI short film and how to storyboard with AI.

FAQ

Can I use ChatGPT to make videos?

You can use it to plan them. You cannot use it, on its own, to generate a video file. If a clip appears in the chat, a third-party app made it.

Can you make videos with ChatGPT?

You can write the script, the shot list and the prompts in ChatGPT, then generate the shots in a video studio. ChatGPT itself does not render the video.

Can ChatGPT generate video for free?

ChatGPT’s Free plan can still write a script and make still images, but it cannot generate a video file, and neither can the paid ChatGPT plans: none of them lists video generation as a feature. Third-party apps inside the chat have their own fees. In CinemaDrop, Lite is free and includes script, storyboard and stills, but video starts on Basic.

Can ChatGPT create videos for free?

It cannot create a video file on any plan, free or paid. Free ChatGPT is still useful for the writing and the stills.

Can ChatGPT generate cinematic videos?

No. It can describe a cinematic shot and draft the prompt. Rendering a cinematic clip takes a video model in a studio. For a model-by-stage view, see which AI is best for filmmaking; on CinemaDrop, the featured options are Kling 4.0 and Seedance 2.5.

Can AI turn my photo into a video?

A dedicated video model can start from a still and generate a clip. ChatGPT can prepare the still, and it may be able to make a looping GIF once that feature has rolled out to your account, but that is not image-to-video.

Can AI generate videos from text?

Yes, in a text-to-video model. ChatGPT can write the text. It cannot run that model itself.

Is it legal to create AI videos?

The legal picture depends on what you generate and how you use it. As a starting point, disclose AI-generated work where you should, and do not copy real people or copyrighted characters. For a fuller discussion, see whether it is legal to make a movie with AI. This is not legal advice.

Which AI is best for producing videos?

It depends on the stage: writing, stills, or generating the shots. Our filmmaking model guide walks through that choice. If you need a free starting point, Lite on CinemaDrop covers script and storyboard, and which AI video generator is free compares free video options.

Which ChatGPT is best for making videos?

There isn’t a ChatGPT video model to choose. Free versus Plus is about usage limits, image tools and apps, not a built-in renderer. For video, you leave ChatGPT and use a studio.

Is Sora still available? Is Sora discontinued?

The Sora web and app experiences were discontinued on April 26, 2026. The Sora 2 models and Videos API were shut down on September 24, 2026, with no one-to-one replacement. OpenAI’s help article and API guide, linked above, are the current sources.

What happened to Sora?

OpenAI discontinued the consumer product in April 2026 and shut down the API in September 2026. Its help article does not state a reason.

Start with the script, then storyboard the shots

So, can ChatGPT generate video? Not as a file, and not with a built-in OpenAI video model. It can still do the writing, the stills and, as the feature rolls out, a looping GIF. For shots you can cut together, take the plan into a storyboard and generate them there.

When you are ready, start a script in CinemaDrop, then storyboard and generate the shots. Plans and credits are on the pricing page.

Published on by Leila Morgan

Leila Morgan

AI Filmmaking Writer
Make your own movie with AI.No experience needed.
Start creating free