Skip to main content

HTML to Video

Render animated app demos, text-message conversations, motion graphics and animated charts as MP4 videos from Claude, through the Arcads MCP.

Written by Nilo

HTML to Video turns an animation written as a web page into an MP4 video, with its sound. Ask Claude (or another AI assistant connected to the Arcads MCP) for an app demo rebuilt from screen recordings, a text-message conversation, motion graphics, kinetic text or an animated chart, and it writes the animation and renders it for you. Because every frame comes from code, text, layout and timing come out exactly as written.
​

It comes with a companion tool, HTML to Image, which turns the same kind of page into still images: to check frames before a video is rendered, or to make a static visual with exact text, for example a headline on a static ad.
​

Who can use it

All plans. You need the Arcads MCP connected to your AI assistant.
​

Where to find it

HTML to Video and HTML to Image are tools of the Arcads MCP. Connect the Arcads MCP to Claude (Claude > Settings > Connectors), then ask for the video in the chat.
​

How to use it

  1. Connect the Arcads MCP to your AI assistant if it isn't connected yet (see Related below).

  2. Describe the video you want: for example "an animated app demo of our checkout flow, 15 seconds, vertical", or "a text-message conversation between two friends about our product". Add any screenshots, logos or product images it should show.

  3. The assistant writes the animation and checks a few frames for free with HTML to Image. Ask it to show you the frames and to change anything you want.

  4. When the frames look right, the assistant renders the video with HTML to Video. It is usually ready within a minute.

  5. Open the video from the link in the chat, or find it in Arcads with your other MCP generations.

Options

  • Length: up to 3 minutes.

  • Size: any shape up to 1080p (1920 px on the long edge, 1080 px on the short edge). Vertical 1080 × 1920 is the default.

  • Sound: voice-over, music and sound effects can be mixed into the same render, each placed at the moment it should play, with its own volume.

  • Images: up to 20 images (screenshots, logos, product shots) can appear in the animation.

  • Timing to a voice-over: the animation can be retimed so each scene lands on the words that describe it.

  • HTML to Image: returns frames in the chat to check, or saves them full-size as images in Arcads.

Credits

  • HTML to Video: 5 credits per started 30 seconds of video. A 15-second video costs 5 credits, a 45-second video 10 credits, a 3-minute video 30 credits.

  • HTML to Image: free, whether the images are only previewed or saved.

  • A render that fails is refunded.

Credit costs here are the standard rates. The Generate button in the app shows what each generation costs on your account.

Where it works

  • MCP: yes (HTML to Video and HTML to Image).

  • Studio: not yet.

  • Mark: not yet.

  • Workflows: not yet.

Limitations

  • Videos are rendered at 30 frames per second, up to 3 minutes and up to 1080p.

  • The animation is drawn from code: it does not generate actors, people or realistic footage. Use the video models or Talking Actors for those, and combine the results.

  • Fonts, icons and animation libraries have to be loaded by the page itself; a font that isn't loaded won't show.

  • Free image renders are limited per workspace per day.

Related

Did this answer your question?