MCP servers that can navigate a browser
2 verified servers expose a tool that can navigate a browser to a URL
Driving a browser starts with going somewhere: open a URL, load the page, work from what renders. A navigate tool is the first call in almost every browser-automation flow, the move that puts the agent on the page it needs to act on.
These verified servers let an agent navigate a browser to a URL.
Playwright
Microsoft
Microsoft's official browser-automation server that drives pages via the accessibility tree, not pixels.
browser_navigatebrowser_navigate_back
browser_navigate drives the browser to a URL and browser_navigate_back steps through history, on a Playwright browser you control for testing or authenticated work.
Browserbase
Browserbase
Cloud-hosted browser automation with Stagehand, so agents drive headless browsers without local infra.
navigate
navigate points a cloud browser at a URL, so an agent gets a clean, disposable session without running a browser of its own.
What to know
Navigation is step one, and what follows is the rest of the automation: snapshot the page, click an element, fill a form, read the result. The two servers here split on where the browser runs. Playwright drives a browser you control, local or in your own environment, which fits an agent testing your app or working behind your authentication. Browserbase runs the browser in the cloud, which suits agents that need a clean, disposable session without managing a browser themselves.
A browser session holds state, cookies, a login, a place in a flow, that vanishes when it ends. An agent that navigates, does half a task, and starts cold next session repeats the login and the steps it already took. What it learned on the page is worth keeping even though the session is not.
Questions
- Local browser or cloud browser?
- Playwright drives one you control, right for testing your own app or working behind your authentication. Browserbase runs the browser in the cloud, which suits a clean, throwaway session an agent does not have to manage. Pick by whether the agent needs your environment or just any browser.
- What does an agent do after it navigates?
- Navigation is the first step. From there the same automation servers snapshot the page, click elements, fill forms, and read results, so an agent can drive a flow end to end. This page covers the open-the-page step every browser task starts with.