Guides

How to make videos in ChatGPT now that Sora is gone (2026)

· 7 min read

Short answer: ChatGPT can no longer make a video by itself. OpenAI closed the Sora app and website on April 26, 2026 and the Sora API on September 24, 2026. To get a video from inside ChatGPT today you install a video plugin from ChatGPT's plugin directory, or you connect a video service as a custom app. The video is then made by that service, under your account with it.

Disclosure: Dora, mentioned below, is made by us, the team behind this blog. Dora is not in the ChatGPT plugin directory yet; we explain the honest way to use it today. Everything about other services comes from OpenAI's help center, OpenAI's public plugin listings and the vendors' own pages, checked on October 7, 2026.

What happened to Sora?

Sora was OpenAI's video model and app. OpenAI announced on March 24, 2026 that it was closing it, and its help center gives the dates:

  • April 26, 2026: the Sora web and app experiences were discontinued.
  • September 24, 2026: the Sora API was discontinued.

OpenAI's help page points anyone with old Sora content to an export at sora.chatgpt.com/sunset, and says Sora data is deleted after any final export window. Many answers online still say "use Sora" because they were written while it existed. That route is closed.

Option 1: a video plugin from ChatGPT's directory

ChatGPT now calls its add-ons plugins. A plugin connects ChatGPT to an outside service; you install it, connect your account with that service if asked, and mention it in a chat with @ followed by its name. OpenAI says the directory is available across ChatGPT plans, while installing or using a given plugin depends on your plan, workspace, role and region. Credits and plans are the vendor's, not OpenAI's.

These video plugins have a public listing on OpenAI's site (openai.com/business/plugins). We grouped them by the kind of video they make:

Plugin What its listing says it does Best for
Runway Generate and edit images, videos and audio using your Runway account Clips from a prompt or an image
Kling AI Image and video generation Clips from a prompt or an image
Higgsfield Images and videos from text or reference images, including animating a still photo Clips from a prompt or a photo
OpenArt Images and video, with reference images and a choice of models Clips from a prompt or a photo
HeyGen AI videos with avatars, voices and translations A presenter talking to camera
VEED Video Generator A library of characters and voices turned into a finished video A presenter talking to camera
invideo Script, visuals, voiceover and music from a text prompt A produced video from a brief
Visla Video Maker Narrated videos from scripts and source material A produced video from a brief
Descript A video editor you direct by chatting Editing a recording you already have

How to install one: open Plugins in the ChatGPT sidebar, search for the name, select Install plugin, and Connect your account with the vendor if it asks. Then, in a chat, type @ and the plugin's name and describe the video. HeyGen's own ChatGPT page, for example, describes typing /heygen in a conversation and getting a production plan to review before it renders.

If a plugin does not appear for you, OpenAI's advice is to check its plan, region and workspace requirements; in a company workspace, an administrator may have to allow it.

Option 2: connect a video service as a custom app

Any service that offers an MCP server (the standard way AI assistants connect to tools) can be added to ChatGPT as a custom app in developer mode, even if it is not in the directory. The catch is the plan. According to OpenAI's help article on developer mode, as of October 2026:

  • Full support, including actions that create things (which is what generating a video is) is available on ChatGPT Business, Enterprise and Edu, rolling out in beta.
  • Pro users can connect custom apps in developer mode with read and fetch permissions only.
  • Custom apps are configured and used on ChatGPT on the web, not the mobile app.
  • In a workspace, admins or owners enable developer mode and publish the app.

So on a personal Plus or Pro account, a custom video app cannot generate for you today. On a Business, Enterprise or Edu workspace it can.

Option 3: Dora, inside ChatGPT or on its own

Dora is our AI video and image generator. It turns a photo or a description into a short video or an image, vertical, horizontal or square. A video takes about two minutes and an image about one, and you see what each generation costs before you start it. It works with a paid Dora account.

Dora is not in the ChatGPT plugin directory yet. Our submission is pending. Until it is listed, there are two honest ways to use it:

  1. As a custom app in ChatGPT, on a plan that allows it (see Option 2). On ChatGPT on the web, open Settings, then Apps, enable developer mode if your workspace requires it, choose to create a custom app, and paste this address:

    https://doravideo.com/mcp

    Sign in to Dora when asked, then ask for a video in normal words, for example: "Animate this product photo as a short vertical clip. Show me the cost and wait for my confirmation." The full steps are on our ChatGPT page.

  2. Directly at doravideo.com or in the Dora iPhone app, with the same account. This works whatever ChatGPT plan you have: write the prompt in ChatGPT if you like, then paste it into Dora.

The same Dora address also works with Claude, Cursor and other assistants that support MCP. For the developer view, see the best video MCP servers for AI agents.

Or skip ChatGPT: standalone video apps

Many people who used Sora simply moved to a dedicated video app instead of a plugin. Common choices are the Kling AI and Runway apps for generated clips, CapCut and PowerDirector when you also want an editor, Viggle for dance and meme templates, and Dora when you want several video models in one place on iPhone and the web. We compare them in the best AI video generator apps for iPhone and the best apps to turn a photo into a video.

Which route should you take?

  • You want to stay inside ChatGPT on a personal plan: install a video plugin from the directory.
  • Your company uses ChatGPT Business, Enterprise or Edu: a plugin, or a custom app such as Dora at https://doravideo.com/mcp.
  • You just want the video: use a video app directly, such as Dora, and use ChatGPT to write the prompt.

Frequently asked questions

Can ChatGPT make videos?

Not on its own any more. OpenAI discontinued the Sora app and website on April 26, 2026 and the Sora API on September 24, 2026, so ChatGPT has no built-in video model. You can still get a video from inside ChatGPT by installing a video plugin from its directory, such as Runway, Kling AI, Higgsfield, HeyGen or invideo, or by connecting a video service as a custom app on a plan that allows it.

Is Sora still available?

No. According to OpenAI's help center, the Sora web and app experiences were discontinued on April 26, 2026 and the Sora API was discontinued on September 24, 2026. OpenAI pointed users to an export page to download their old Sora content.

How do I add a video app to ChatGPT?

Open Plugins in the ChatGPT sidebar, search for the video plugin by name, select Install plugin and connect your account with that service if asked. Then mention it in a chat with @ and its name and describe the video. Whether a plugin is available to you depends on your plan, workspace, role and region.

Which ChatGPT plan do I need to connect a custom video app?

OpenAI says full custom MCP apps, including actions that create content such as generating a video, are available on ChatGPT Business, Enterprise and Edu, on the web. Pro users can connect custom apps in developer mode with read and fetch permissions only. Plugins from the directory are a separate route and depend on each plugin.

Is Dora available in ChatGPT?

Not in the plugin directory yet; Dora's submission is pending. On an eligible ChatGPT plan you can add it today as a custom app with the address https://doravideo.com/mcp, or use it directly at doravideo.com or in the Dora iPhone app with a paid Dora account.

What is the best alternative to Sora for making videos?

It depends on the video. For clips generated from a prompt or a photo, people use Runway, Kling AI, Higgsfield or Dora. For a presenter talking to camera, HeyGen or VEED. For a produced video with voiceover from a script, invideo or Visla. Inside ChatGPT, each of these works through a plugin or a custom app; outside it, through the vendor's own app or website.

Sources

Try it on Dora

Keep reading

ComparisonsThe Best Video MCP Servers for AI Agents (2026 Comparison)Looking to give your AI agent video generation powers? We compare the leading video MCP servers across model access, setup friction, pricing, and output reliability.ComparisonsThe best AI video generator apps for iPhone (2026)Seven iPhone apps that generate video with AI, from text or a photo, ranked. What each does best and how it charges.ComparisonsThe best apps to turn a photo into a video on iPhone (2026)Eight iPhone apps that animate a photo with AI, ranked. What each one is good at, how it charges, and which one to pick.