- IMAGE
- TEXT
- CODE
- HUMAN
Short answer
How can my AI take screenshots of websites or HTML?
Your AI uses Gotenberg to render URLs, HTML snippets, or Markdown into images or structured text. You can grab a full page view or a specific code snippet to feed back into your chat. This allows your agent to see exactly what a user sees on a webpage.
Where every snapshot lands.
Visual and text outcomes
The AI processes your request and delivers the result in one of these formats.
IMAGE
Visual snapshots
The AI renders a URL or HTML block into a PNG. You get a clear picture of the layout or design.
TEXT
Markdown content
The AI converts web content into clean Markdown. This makes it easy for your agent to read and process the text.
CODE
Raw HTML snippets
The AI captures specific HTML blocks. Use this when you need to inspect the underlying structure of a page.
HUMAN
Manual review
The AI provides the file or text. You decide if the capture is accurate enough for your next step.
The workflow
What your AI does when a request arrives.
The AI acts as the bridge between your prompt and the web renderer.
Capture a URL
The agent visits a specific web address and takes a visual picture of the rendered page.
screenshot_urlRender HTML
You provide a block of code, and the AI turns that raw markup into a visual image.
screenshot_htmlConvert to Markdown
The agent pulls the content from a page or snippet and turns it into readable Markdown text.
screenshot_markdownDeliver results
The AI hands the image or text back to you in the chat interface.
screenshot_urlscreenshot_htmlscreenshot_markdown
Try it
Copy these to start.
Use these prompts to trigger the screenshot tools immediately.
Starting points
Swap the bracketed text with your actual URL or code snippet.
Gotenberg Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_IkDIVkylexUE9eZzBPNF2NCeizigqyr4NO7jaNq0/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Gotenberg capabilities are ready to use.
{
"mcpServers": {
"gotenberg-mcp": {
"url": "https://edge.vinkius.com/vk_preview_IkDIVkylexUE9eZzBPNF2NCeizigqyr4NO7jaNq0/mcp"
}
}
}Take a screenshot of https://example.com to show me the layout.
Convert this HTML snippet into an image: <div><h1>Hello World</h1></div>
Turn the content at https://news.ycombinator.com into Markdown so I can read it.
Capture a visual snapshot of the landing page at https://github.com.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Gotenberg once, then ask.
Just click the link to connect the MCP to your client. Your credentials stay encrypted, and you can start using the tools right away.
Connect Gotenberg to your AIFAQ
How the screenshot task behaves
- 01
Can the AI edit the website before taking the screenshot?
No. The AI can only capture the page as it exists or as it is rendered from the HTML you provide.
- 02
Can the AI save these images directly to my hard drive?
No. The AI provides the image or text within your chat interface. You must save it manually if you want it on your disk.
- 03
Can I take screenshots of pages that require a login?
No. The tools capture the public-facing URL or the specific HTML you provide. They cannot bypass authentication walls.
- 04
What is the difference between the HTML and URL tools?
The URL tool visits a live website, while the HTML tool renders a specific block of code you paste into the chat.
- 05
Can the AI interact with buttons on the page?
No. The tools are for capturing snapshots of the current state, not for navigating or clicking through a site.
More questions about Gotenberg? The Connector page answers them. See everything the Gotenberg Connector can do
Connect Gotenberg to Claude, Cursor, ChatGPT & more
