Browser Automation

Automate websites, fetch page info and schedule recurring browser sessions

Automate browsers with the Playwright CLI for navigation, screenshots and E2E tests

Automate websites from a plain-language goal, including logins, forms and extraction

Choose the built-in browser for web pages, logins and dynamic sites

Automate Chrome with Puppeteer for screenshots, performance and JavaScript debugging

Control native desktop apps through accessibility trees with snapshots and clicks

Create cloud browser sessions, fetch pages and track Browserbase usage from the CLI

Click buttons, links and options on a web page by CSS selector or visible text

Read page text and element values from the current browser tab

List open browser tabs with their index, URL and title

Open a URL in the browser and start a web automation session

Open a new browser tab at a URL while keeping the current page open

Launch the browser or check whether it is running and which page is open

Capture a screenshot of the current web page content for review

Switch to a different open browser tab by its index number

Describe a multi-step web task in plain language and let the browser finish it

Type text into web page inputs to fill forms and enter search queries

Click buttons, icons and menu items in desktop apps by name or coordinates

Locate desktop buttons, menus and icons with UI Automation or vision

Send keyboard shortcuts like copy, paste, save and undo to desktop apps

Capture a Windows desktop screenshot and save it as a file to share

Scroll up, down, left or right in a desktop app with the mouse wheel

Type text into the focused desktop app, including Chinese input

Wait for a desktop dialog, window or UI element to appear before continuing

List, switch, minimize, maximize and close desktop windows by title

Run GitHub, YouTube, Bilibili and X commands from the CLI using your Chrome logins

Operate macOS and Windows desktop apps through native accessibility trees

Automate macOS desktop apps, covering permissions, menus, sheets and Notification Center

Automate Windows desktop apps and work around PowerShell quoting and UIPI limits

Run a desktop app from one plain-language goal without loading its full UI tree

Read and drive embedded web terminals such as xterm.js through a browser

Operate an authorized Android phone for app tasks and mirror its screen

Run natural language mobile tasks on an Android phone through OpenGUI

Complete authorized Android tasks and iOS simulator tests from screenshots

Guidance for running browser automation workflows reliably

Navigate websites for research, forms and data extraction with browser patterns

Operate many websites in parallel with up-to-date browser action manuals

Automate browser tasks like form filling, screenshots and page data extraction

Supply verified Actionbook page selectors to Rust documentation research workflows

Drive the agent-browser CLI for crate and docs research that needs a live browser

Drive the Kimi desktop app for data queries, research reports and macOS automation

Research and inspect web pages through a managed browser session

Drive a named macOS app with raw mouse and keyboard input

Control a connected desktop with screenshots, clicks, typing and scrolling

Point at things on screen so the assistant knows what you mean

Browse websites and read pages from the command line using a browser CLI

Add human-like mouse, keyboard and scroll behavior to browser automation

Launch a stealth Chromium build with fingerprint patches to get past anti-bot checks

Automate browser sessions, scrape pages and run browser tasks with Notte