SquadOS SquadOS
EN
Start
Visual rendering in chat

Visual assets generated by the agent, ready to download.

The agent builds the creative's HTML/CSS and calls the tool to render the image in the right format and size, without opening an external editor.

Try it now
In practice

What changes in the operation

Some agents need to deliver more than text: a promotion card, announcement banner, square social post, or visual campaign summary. With HTML to Image, the agent creates the HTML/CSS and the tool renders the final file in the conversation. You standardize dimensions and format per agent, and the user receives the asset ready to download or send.

Ready-to-use formats

Render as webp, png, or jpeg, with a per-agent default so users do not choose the format every time.

Dimensions by channel

Set width and height in pixels for feed, story, banner, thumbnail, or any operational format.

Full HTML and CSS as input

The tool receives full HTML with inline or head-embedded styles, enabling predictable and reusable layouts.

Per-agent defaults

Configure default format, width, and height so the agent follows the channel identity without repeating parameters.

HTML to Image

How it works

The agent composes the layout. The tool delivers the final image.

The user asks for the asset

Example: “Create a 1080x1080 card to announce Friday's promotion.”

The agent builds the HTML/CSS

It writes the layout, applies the content, and calls the tool with html, width, height, and format.

The image returns ready in the conversation

The tool renders the file in the chosen format and returns the result for download, review, or sending.

Frequently asked questions

  • What output formats are supported?

    The tool supports webp, png, and jpeg. You can set the default in the agent settings or specify it per call.

  • Can I set a fixed size for all creatives from this agent?

    Yes. In the agent settings, you define the default width, height, and format for the HTML to Image tool. The agent uses those values when there's no explicit specification in the request.

  • Does the agent need a special prompt to generate creatives?

    Not necessarily. Just enable the HTML to Image tool on the agent and describe in the prompt that it should generate visual creatives in HTML. AgentMaker can do this for you automatically.

  • Does external CSS (via link or @import) work?

    The tool processes full HTML with inline or <head>-embedded styles. For maximum compatibility, prefer inline or embedded styles — external dependencies may not load in the rendering environment.

Give the agent a visual output, not only text

Enable HTML to Image to generate cards, banners, and simple assets inside the conversation flow.

Start for free

webp · png · jpeg. Customizable dimensions.