Browser Automation API

Give it a URL; it opens the page in a real browser, lets the JavaScript run, and hands back the title, the visible text and every link. Delivered into Claude and other AI assistants over MCP — nothing to install, no browser on your machine.

Why this exists

A plain HTTP fetch downloads the HTML a server sends and stops there. Most modern sites build their navigation and body after that, in the browser. So a fetch of a real page often returns a shell with almost nothing readable in it.

A live example, both run through this tool on the same URL — the National Funeral Directors Association media page:

• Rendered in a browser: 4,338 characters of text, 100 links.
• Free preview of the same render: the real page title and the first 500 characters, links withheld.

Try it before you pay

Connect an MCP-capable assistant to product.kcalbinmcp.uk/mcp and call browser_fetch immediately. No account, no card. You get the real page title and the first 500 characters of the real render, so you can confirm it actually works on your pages before paying. The free preview is limited to 10 renders per day.

Plans

Starter — $29/month

• 500 page renders per day
• Full page text, up to 20,000 characters
• Every link on the page, with its anchor text
• Same API key as any other Kcalbin MCP product you subscribe to

Subscribe — $29/mo

Pro — $99/month

• 2,500 page renders per day
• Everything in Starter

Subscribe — $99/mo

What it does not do

Stated plainly, because the gap between this and a dedicated scraping platform is real and you should know it before you pay:

• No proxy rotation and no residential IPs. Requests come from one datacenter address, identified honestly as KcalbinBrowserAPI/1.0.
• No anti-bot bypass. A site running commercial bot protection will block it, and we do not try to defeat that.
• No geotargeting — you cannot choose the country a request comes from.
• Read-only. It renders a page and extracts from it. It does not click, fill forms, log in, or take screenshots.
• Public pages only. Private, internal and cloud-metadata addresses are refused outright, and so is any hostname that resolves to one.

If you need rotating residential proxies, anti-bot bypass or millions of renders a month, buy a dedicated scraping platform — that is what they are for, and they are good at it. This is for the case where you want a page read properly from inside the assistant you are already working in, billed on the key you already have.

Fair use

Daily quotas are real limits, not guidance, and they reflect measured capacity on the hardware that serves them. Use the tool on pages you are permitted to access, and respect the terms and robots policy of the sites you point it at.