Works with Claude, ChatGPT, Cursor

Ask in chat.
Get a video.

Connect VideoGenAPI once, then describe a scene the way you would to a colleague. Claude or ChatGPT handles the rest: Sora 2, Veo 3, Kling 3 and more, billed on the plan you already have.

Connect in two minutes Create a free account
Claude ChatGPT Cursor Claude Code OpenClaw Hermes

Stay in the conversation

No extra app to learn. You already chat with an assistant, and now it can actually produce the video.

chat

Plain-language prompts

Talk about lighting, camera and mood. You do not need to remember endpoint names.

movie

The models you already pay for

Sora, Veo, Kling and the rest of the VideoGenAPI catalog. One subscription, from chat or from the API.

bolt

Ready in a couple of minutes

Most clips land between thirty seconds and a few minutes. The assistant shares a link when it’s done.

Connect your assistant

Choose the app you use. We show the three short steps for that one.

Add VideoGenAPI in Claude

Claude opens a sign-in window. No keys to copy.

  1. 1 In Claude, go to Customize → Connectors and add a custom connector.
  2. 2 Call it VideoGenAPI and paste the link below.
  3. 3 Sign in, then ask Claude to make a video.
https://videogenapi.com/mcp
Open Claude

What people ask for

A sentence is enough. Here are prompts that work well on day one.

Social

A vertical product unboxing

Handheld, captions, bathroom or kitchen light, the kind of clip that belongs on Reels or TikTok.

Brand

A quiet cinematic establishing shot

Golden hour, slow camera, 16:9. Ask for Sora 2 or Veo 3 when you want it to feel like a film.

Still → motion

Bring a photo to life

Drop an image link in the chat and say how it should move: a slow orbit, wind in fabric, steam over a cup.

Ads

A five-second hook

Cold open, one visual surprise, then the benefit. Fast models like Grok Imagine or Seedance are enough to iterate.

Name a model, or don’t

Say “use Veo” if you care. Otherwise the assistant picks something sensible from your catalog.

Sora 2

Story-like shots that hold together from start to finish.

Veo 3

Photoreal scenes when you want it to look shot, not generated.

Kling 3

Product motion and social formats, including from a still.

Grok Imagine

Quick drafts while you are still finding the idea.

Plus Seedance, Wan, LTXV and the rest of the model list.

Questions, quickly

What is this, in one sentence?

add

It lets Claude, ChatGPT, Cursor and similar apps create videos with your VideoGenAPI account, using the same models and plan as the website.

Do I need to know how to code?

add

Not for Claude or ChatGPT. You paste a link, sign in, and ask for a video in plain language. Cursor and Claude Code need a short config if you already use those tools.

Will this cost extra?

add

No. Videos use your VideoGenAPI subscription. Connecting is free. You need an active plan to generate, just like on the API.

Which models can I use from chat?

add

The same ones as on VideoGenAPI: Sora 2, Veo 3, Kling 3, Grok Imagine, Seedance and the rest of the catalog. Name the model in your prompt, or let the assistant pick.

How long until I get a video?

add

Usually between thirty seconds and a few minutes, depending on the model. The assistant waits, then shares a link when it is ready.

Is it safe to connect Claude to my account?

add

You approve access on a sign-in screen. The assistant can create videos and see your usage. It cannot change your password or billing.

Prefer the written spec? It’s on the MCP docs page.

Try it on the next prompt

Connect Claude or ChatGPT, then ask for the video you were going to brief anyway.

Connect VideoGenAPI See plans