MCP server · Browser automation
Web browsing MCP server
by kimtth
Your AI can open pages in a real browser, read what is on them, and take screenshots for you.
This is a small helper that lets your AI drive a real web browser on your own computer. You type a request in plain words, and the browser it controls goes to the page, looks at it, and brings something back to the chat. It is handy for people who want their AI to check a live web page instead of guessing, and it is also used as a teaching example for how these helpers are built.
What is an MCP server? The 30-second version
On its own, your AI can only chat with you. It cannot open a web page or click anything. An MCP server is a small helper program that gives your AI one extra skill, in this case the skill of browsing the web, so your AI can go and look something up for you when you ask. You stay in the chat. The helper does the clicking in the background.
What this MCP server does
You ask your AI something that needs a real web page, like a page address you want it to visit. Your AI passes that request to this helper program. The helper opens a Chromium browser window on your computer, goes to the page, and keeps that page open so it can look again if you ask a follow-up question. It can read the text on the page or take a picture of it. Then it hands the result back to your AI, which puts the answer in the chat for you.
Click to zoomWhat you can do with it
- Open a web page by its address
- Read the text on a page so your AI can summarize it
- Take a screenshot of a page
- Keep the same page open so you can ask follow-up questions about it
- Grab a specific piece of a page, such as a selected element
- Check what a page looks like right now rather than trusting old information
Try asking your AI
- “Open https://example.com and tell me what this page is about”
- “Go to our company news page and summarize the latest three headlines”
- “Take a screenshot of the page you just opened”
- “Open this page and pull out the price shown at the top”
What it gives back to you
Mostly you get words back: a summary of the page, the text it read, or an answer built from what it saw. When it takes a screenshot, you get that image in the chat. It can also hand back a specific piece of the page, like the text of one element you asked about. Everything shows up in your normal chat, so it looks like your AI simply answered you.
Before you start
What you need
- A Windows, Mac, or Linux computer with a normal desktop screen (the browser window is visible)
- Python 3.13 or newer, plus the Tkinter part of Python
- The uv tool, which is used to run the project (the README links to a free install guide)
- An Azure OpenAI account with a model deployment that supports tool calling, and its address and key
- About 10 minutes of setup, including downloading the Chromium browser the helper uses
Good to know
The helper opens a real browser that can visit any page you name and reads what is on it, so avoid pointing it at private pages you would not want fetched, and keep your Azure key in the local .env file only.
Install it with your AI
Add Web browsing MCP server to your AI, no technical skills needed
You don't install anything by hand. You copy one prompt, paste it into an AI that can work on your computer, and it checks, installs and connects the server for you, asking you when it needs something.
Sign in to get the install prompt
Members get a ready-made prompt that lets the Claude desktop app check Web browsing MCP server, install it and connect it for them, step by step. You don't need any technical skills: you copy, paste and answer a few questions. Your connected AI can also find and install any of the 4,066 MCP servers here for you.
Who it's for
People who want their AI to look at real web pages, and learners who want a small, readable example of how an AI browsing helper is put together.
