Can ChatGPT Make Videos Now That Sora Is Gone?

Sora is closed, so ChatGPT makes no video on its own. What works now: plugins inside the chat, which plugin fits which video, and the steps to make one today.

Per Harald Borgen

Per Harald Borgen

— 11 min read
Can ChatGPT Make Videos Now That Sora Is Gone?

Not on its own, and yes through a plugin. OpenAI closed Sora's web and app experiences on April 26, 2026 and the Sora API on September 24, 2026, so ChatGPT no longer has a video model of its own. What it has instead is a directory of plugins: you install one, mention it in a chat, and a vendor's service turns the conversation, a document or a prompt into a video.

Last verified: October 23, 2026. Checked for this article: OpenAI's Sora discontinuation and plugin help articles, OpenAI's DevDay 2026 recap, the plugin listings in ChatGPT and on openai.com, the vendor pages named below, and Scrimba's Explain documentation.

Scrimba publishes this article and makes one of the plugins in it: Scrimba Explain, listed in ChatGPT as Explain Video Generator. It is named where it fits, and the tools that fit other jobs better are named too.

What Happened to Sora?

Sora was OpenAI's text-to-video model, and it is gone: first the apps, then the developer API, so no current ChatGPT plan includes an OpenAI video generator.

OpenAI's help center states the first date in one line:

"The Sora web and app experiences were discontinued on April 26, 2026."

The same article gives September 24, 2026 as the end date for the API, a date that has now passed. Anyone with Sora content was pointed to an export at sora.chatgpt.com/sunset, and purchased ChatGPT or Sora credits can still be used for Codex. Many answers still say "use Sora" because most pages about ChatGPT video were written while it existed.

How Does ChatGPT Make a Video Now?

A ChatGPT plugin is an installable package that connects ChatGPT to an outside service, and video plugins hand the actual rendering to that vendor. ChatGPT writes and steers; the vendor's model or renderer makes the file.

OpenAI's plugins help article describes the flow: open Plugins, select a plugin, choose Install plugin, and Connect a vendor account if the plugin asks for one. Then you call it with an @ mention, or ChatGPT uses an installed plugin on its own when it fits. Availability varies:

"The plugin directory is available across ChatGPT plans. Installing or using an individual plugin depends on your plan, workspace, role, region, and its included capabilities."

In practice, ChatGPT video means somebody else's video, ordered from inside ChatGPT, usually with the vendor's account and credits. Runway's listing says its plugin works "using their Runway account" and that "purchases are completed on runway.com, never inside the chat." HeyGen's ChatGPT page says you need a free HeyGen account and that generation uses HeyGen credits.

Explain asks for less up front. Its listing says "Free to start, no login required": no Scrimba login for a first explainer, though plugins still need a signed-in ChatGPT session.

What changed on September 29, 2026. At DevDay, OpenAI announced plugin extensions, which give a plugin "a home in the sidebar" and panels beside the chat, and wrote: "Improved ranking and recommendations help people find relevant plugins in the directory and in conversations." Two plugins that make video arrived around it. Prezi's plugin, announced for educators on September 25, turns slides into an interactive deck that its listing says you can share and export "as a narrated video or a podcast". Pika's plugin followed on October 16; its listing offers product ads, short films, photo animation and lip-sync. For a plugin-by-plugin comparison with prices, see Scrimba's guide to the best ChatGPT plugins for creating videos.

Which Tool Makes Which Kind of Video?

The right tool depends on the video you want: a generated clip, a presenter, produced footage or an explainer. A plugin that is good at one is usually weak at the others.

Checked on each vendor's own page or plugin listing on October 23, 2026:

The video you want ChatGPT alone? The route that works now Someone on screen? What you need
A script, storyboard or shot list Yes Ask ChatGPT directly No video yet Your ChatGPT plan
A cinematic clip from a prompt No Runway or Pika plugin in ChatGPT; outside ChatGPT, Gemini Omni in the Gemini app Only if generated A Runway or Pika account and credits; a Google AI subscription for Gemini Omni
A talking presenter or avatar No HeyGen, VEED Video Generator or Synthesia plugin Yes, an AI avatar A free HeyGen account, with video paid in HeyGen credits; a paid Synthesia account for Synthesia
A produced video with voiceover, footage and music from a script No Visla Video Maker or invideo plugin Stock or generated footage A Visla account, for Visla
A narrated version of a slide deck No Prezi plugin Only if you record yourself A Prezi account
A narrated explainer from your notes, a PDF, a web link or the chat No Explain Video Generator plugin (Scrimba Explain) No, typed slides with narration A signed-in ChatGPT session; free to start
An edit of a recording you already have No Descript plugin Whatever you filmed Install it and connect when asked

The details behind the rows: Google says Gemini Omni will replace Veo in the Gemini app; it makes 10-second videos with native audio and requires a Google AI subscription. VEED's listing reads "Pick your character. Pick your voice. Generate your video." Synthesia's plugin, in beta, builds a video from a prompt, a script, a document or a link and requires a paid Synthesia account; Synthesia's free plan, with 10 minutes of video a month, is used on its own site.

Visla's listing says it builds narrated videos "with stock footage and background music" and that "A Visla account is required." The invideo listing promises script, visuals, voiceovers and music from text prompts. Prezi needs a Prezi sign-in and offers Newscaster mode if you want to present on camera. Descript's listing sums it up as "Edit video by chatting."

What Does an Explainer Made in ChatGPT Look Like?

An explainer is a narrated lesson built from typed slides, diagrams and images, not a filmed or generated movie.

Scrimba Explain, listed in ChatGPT as Explain Video Generator with the tagline "Free AI explainer video maker", is one of four routes into the product, shown as cards on Scrimba's hub: the ChatGPT plugin, a Chrome extension, Claude and MCP for codebases, and an API that is in beta, on request. On scrimba.com itself you can attach a PDF or images; save a Word file as PDF first.

The

The listing says it turns "a topic, a document, a webpage, a codebase or a product brief" into a video "with voiceover, diagrams, animations and images." Scrimba's introduction is precise about the format: "An explainer is not a screen recording. Each slide is real, typed content." Explainers are narrated in 33 languages, and they stream, so the first slide appears within seconds and the rest builds while you watch.

In ChatGPT, the model you are chatting with writes the slides, the narration and the diagrams, so a stronger model gives a better explainer. According to the ChatGPT docs page, the plugin then "replies with an explainer card in the chat", and Watch explainer on the card opens it on scrimba.com. A screenshot you uploaded, or an image ChatGPT generated earlier in the chat, can go on a slide; pictures uploaded from the ChatGPT mobile app cannot. The explainer is square unless you ask for a vertical video.

A slide from "What is a RESTful API?", a published example in the use-case gallery:

A slide from the published explainer

For medical students and nursing students, the Healthcare example "Cardiac Prescription Logic" shows formulas as real notation, not as plain text like x^2. Scrimba's guides to turning study notes into a video in ChatGPT and to explaining a pull request to your team cover those two routes, and an architecture onboarding video works the same way from a design doc.

A slide titled

There is no presenter and no camera: nobody is filmed, no avatar speaks, and captions run under the narration. If your video needs a face, use the table above.

How to Make a Video in ChatGPT, Step by Step

To make a video in ChatGPT, install a video plugin, mention it with your material, then keep and check what it returns. The short version below uses Explain Video Generator; Scrimba's guide How to Make a Training Video in ChatGPT walks the same route with a real SOP, prompt by prompt.

  1. Install the plugin while signed in to ChatGPT. Open the listing and choose Install plugin. Scrimba's walkthrough on how to install a plugin in ChatGPT covers each screen.
  2. Mention it with your material. Type @, pick Explain Video Generator, and attach a PDF, paste a link, or ask it to turn the conversation into a video.
  3. Keep what it makes. It starts unlisted: watchable by link, not listed or searchable. Choose Connect Scrimba when ChatGPT offers it, and every explainer is saved to your Scrimba account as it is created. Without the connection, open it and click Save to profile.
  4. Check it before you share or download it. Every explainer page carries this line:
"Explain is AI and can make mistakes. Please do your own diligence."
Scrimba's docs section

On cost, Explain is free to start: Scrimba's plans page allows 1 explainer without an account, 5 on a free account and 100 a month on Pro. An explainer counts when it lands in your account, so with Scrimba connected, each one counts the moment it is made. A free account also gets 3 edits per explainer and 1 exported explainer, and its slides show the Explain logo, which only Pro switches off. Downloads are MP4, 1080p at 60 fps by default, with a .vtt subtitles file.

Before you paste company, client or personal material into a chat, read the data notice on the listing: "When connected to Explain Video Generator, ChatGPT may share relevant chats and memories with this app to help provide context for your requests."

What This Means for Trainers, Teachers and Lecturers

For these three groups, ChatGPT video mostly means narrated explainers from material they already have, each with a different limit.

Trainers

The listing's own example card is "Turn this employee onboarding PDF into an explainer video", and its use-case list includes "Onboard new employees without filming a thing." It makes no SCORM package, tracks no completions and shows no avatar, so the usual route is an exported MP4 uploaded into your LMS, which does the tracking. Scrimba's plans page lists commercial use of explainers, and unlimited exports, under Pro. Read the data notice above before attaching an internal handbook in ChatGPT.

Teachers

The listing names a specific use: "Start your class with a 3-minute lesson opener built from the textbook chapter." The catch: quizzes and the seven teaching styles exist only on scrimba.com, so a plugin explainer has neither. Scrimba's guide to a lesson opener video covers the classroom version.

Lecturers

The card "Make an explainer video from my lecture notes" fits a weekly recap. Slides go in as a PDF, since no PowerPoint upload is documented (turning a slide deck into a video covers the export). No length limit is documented either, so split a long deck into a playlist and check each slide. Scrimba's guide to lecture recap videos covers the weekly routine.

When Another Tool Fits Better

Explain is the wrong tool when the video needs cinematic footage, a face or tracking. Go here instead:

  • A shot that never existed, such as a drone sweep over a city: Runway or Pika inside ChatGPT, or Gemini Omni outside it.
  • The same presenter in every video: an avatar tool. HeyGen, VEED and Synthesia have ChatGPT plugins; Synthesia's needs a paid Synthesia account. Explain has no avatars.
  • Completion records for compliance training: an LMS or a SCORM authoring tool; Explain counts only views and likes.
  • Footage you already filmed: Descript, which edits recordings by chat.
  • A 200-page manual covered in one video: no length limit is documented, so split the manual by section into a playlist and check each lesson.

Frequently Asked Questions

Can ChatGPT make videos?

Not by itself since Sora closed. OpenAI discontinued the Sora web and app experiences on April 26, 2026 and the Sora API on September 24, 2026. ChatGPT makes videos through plugins instead, such as Runway or Pika for generated clips, HeyGen for avatars and Explain Video Generator for narrated explainers.

How do I make a video with ChatGPT?

Install a video plugin while signed in, then type @ and its name with your material: a shot description for Runway or Pika, a script for Visla Video Maker or invideo, or notes, a PDF or the conversation itself for Explain Video Generator. The vendor renders the video, so keep it in the vendor's account and check it before you share it.

Is there a ChatGPT video generator?

Not one made by OpenAI. A ChatGPT video generator is now a plugin from the directory. The one named that way is Explain Video Generator (Scrimba Explain), which makes narrated explainers; Runway and Pika generate clips, and HeyGen, VEED and Synthesia put an avatar on screen.

Is Sora still available in ChatGPT?

No. The Sora web and app experiences closed on April 26, 2026, and the Sora API followed on September 24, 2026. No current ChatGPT plan includes an OpenAI video model.

Can you make videos in ChatGPT for free?

Partly. The plugin directory is available across ChatGPT plans, but each plugin has its own terms. Explain is free to start: 1 explainer without a Scrimba account, 5 on a free account, 100 a month on Pro. A free account exports 1 explainer and shows the Explain logo on its slides. HeyGen gives new accounts free credits to start.

Can ChatGPT make a video with a person talking?

Only through an avatar plugin: HeyGen, VEED Video Generator, which lets you pick a character and a voice, or Synthesia, which needs a paid Synthesia account. Explain Video Generator has no presenter or avatar; it makes narrated slides, diagrams and images.

Can I download a video made in ChatGPT as an MP4?

With Explain, yes, once the explainer is in your Scrimba account, through Connect Scrimba or Save to profile. Download gives an MP4, 1080p at 60 fps by default, with a .vtt subtitles file. The explainer must be public or unlisted, and a free account can export one explainer; Pro exports are unlimited.

Does it work in the ChatGPT mobile app?

Yes. Scrimba's documentation says plugins work in ChatGPT on web, desktop and mobile, and in Codex in the ChatGPT desktop app. One catch: pictures uploaded from the ChatGPT mobile app cannot be fetched, so an explainer made there is built without them.

Key Takeaways

  • ChatGPT has no video model of its own since Sora closed: web and app on April 26, 2026, API on September 24, 2026.
  • ChatGPT makes video through plugins: you install and @ mention them, and the vendor renders the result.
  • Pick by job: Runway or Pika for clips, HeyGen, VEED or Synthesia for avatars, Visla or invideo for produced videos, Explain Video Generator for narrated explainers.
  • Explain is free to start: 1 explainer without an account, 5 on a free account, 100 a month on Pro. With Scrimba connected in ChatGPT, each explainer is saved and counted as it is made.
  • Check every AI video against its source before you share it, because every tool here can get facts wrong.

Sources

All sources accessed October 23, 2026.

Read more