html2img - HTML to Image API

Give your AI agent production-quality rendering: this MCP server turns the HTML your agent writes into pixel-perfect PNG images and A4 vector PDFs, and captures live web pages by URL. It is the official MCP server for the [HTML to Image API](https://html2img.com/), rendered on real headless Chrome, with every result uploaded to a fast CDN and returned as a shareable URL. ## Tools - **generate-image-from-html** - render HTML and CSS into a PNG (social cards, OG images, charts, banners) or a paginated A4 PDF with selectable text (invoices, reports, certificates). PNG results include an inline preview, so the agent can see what it made and refine the markup. - **screenshot-url** - capture any public web page as a full-page or viewport PNG screenshot, or archive it as a vector PDF. Supports custom viewports, element selectors, injected CSS, waits and retina scaling. The same rendering engine powers our [Screenshot API](https://html2img.com/screenshot-api/) and [HTML to PDF API](https://html2img.com/html-to-pdf/) if you would rather integrate directly over REST. ## Authentication You will need an API key (`htim_...`) from a paid HTML to Image plan - each render costs one credit. Sign up at [html2img.com](https://html2img.com/pricing/), create a key in the dashboard, and enter it when connecting. Setup guides for Claude, Cursor and other clients are on the [MCP page](https://html2img.com/mcp/) and in the [installation docs](https://html2img.com/docs/mcp/installation/).

接入信息

传输形态

http

鉴权方式

鉴权未知

端点

https://html2img.run.tools

{
  "mcpServers": {
    "html2img - HTML to Image API": {
      "url": "https://html2img.run.tools"
    }
  }
}

能力清单