Selenium
The long-running open-source WebDriver standard for driving real browsers, now adding WebDriver BiDi for event-driven control.
Quick decision
What it does
Selenium is a browser automation framework and ecosystem built around WebDriver, which drives a browser natively, the way a user would, either locally or on a remote machine through the Selenium server. The project describes WebDriver as both the language bindings and the implementations of the individual browser controlling code, and Selenium WebDriver is a W3C Recommendation.
For agent builders, Selenium matters because it is the cross-browser baseline: the same test-style scripts run against Chrome, Firefox, Edge, and Safari through vendor drivers, and Selenium Grid scales those sessions across machines. WebDriver BiDi, the W3C standard bidirectional protocol created by the Selenium project together with browser vendors, adds a WebSocket connection so scripts can stream and react to browser events such as network requests, console messages, and JavaScript errors.
The trade-offs are operational. You manage browser and driver versions, Grid infrastructure, waits, and flakiness yourself, and the programming model is imperative test automation rather than agent-native planning. Teams that need an agent to decide what to click usually pair Selenium with a higher-level agent layer.
Verified capabilities
Core capability
WebDriver browser control Source verified
Drives a browser natively as a user would, locally or on a remote machine using the Selenium server.
Language bindings and drivers Source verified
WebDriver covers both the language bindings and the implementations of the individual browser controlling code for each browser.
Web elements Source verified
Identifies and works with element objects in the DOM for reading and interacting with page content.
Actions API Source verified
A low-level interface for providing virtualized device input actions to the web browser, such as keyboard and pointer sequences.
WebDriver BiDi Source verified
W3C standard bidirectional protocol with a WebSocket connection for streaming browser events like network requests, console messages, and JavaScript errors, positioned as the cross-browser replacement for the Chrome DevTools Protocol.
Support classes Source verified
Optional higher level features built on the core WebDriver API for common automation tasks.
Quick start
First Python WebDriver script
Selenium 4 provisions the matching browser driver automatically through Selenium Manager.
pip install selenium
from selenium import webdriver
driver = webdriver.Chrome()
driver.get("https://example.com")
print(driver.title)
driver.quit()Source: https://www.selenium.dev/documentation/webdriver/. Examples use placeholders only. Never paste a real key into a profile, config file you share, or a ticket.
MCP support: Selenium is not itself an MCP server. Agents use it through code, and community MCP wrappers exist outside the project.
Works with
Frameworks
LangChain, CrewAI, Any agent able to call a Python or Java function
SDKs and packages
Python, Java, JavaScript, C#, Ruby
Works with
Selenium Grid, Browser drivers from each browser vendor
Only sourced support is listed. A missing framework means AgentsUse has not verified it yet, not that it cannot work.
Health and maintenance
34,522
Checked 2026-10-07
8,718
Checked 2026-10-07
GitHub API repository counts and Apache-2.0 license for SeleniumHQ/selenium, fetched 2026-10-07.
Pricing and license
Open source under Apache-2.0. Hosting browsers, Grid nodes, and any cloud browser grid is your own cost.
Limitations and safety
- You own browser and driver lifecycle management, plus Grid operations if you scale out.
- The model is imperative scripting, so agent planning and recovery need a layer you build or adopt.
- Waits, timing, and flaky selectors remain manual work at scale.
Browser and data tools can read pages, fill forms, and download files. Start with a test account or read-only access, keep credentials in environment variables, and review agent actions before connecting anything that can spend money, send messages, or delete data.
Alternatives to Selenium
Playwright
Modern cross-browser automation with auto-waiting and first-party agent tooling
Tradeoff: Its profile lists its own limits, pricing, and verification details for a side-by-side read.
Puppeteer
Lighter Chrome-focused control with deep Chrome DevTools Protocol access
Tradeoff: Its profile lists its own limits, pricing, and verification details for a side-by-side read.
Browser Use
Agent-native browser control where the model plans steps
Tradeoff: Its profile lists its own limits, pricing, and verification details for a side-by-side read.
Common questions
What does Selenium do for an AI agent?
The long-running open-source WebDriver standard for driving real browsers, now adding WebDriver BiDi for event-driven control.
Is Selenium open source?
Yes. This profile records the license as Apache-2.0 from the official repository.
Does Selenium support MCP?
Selenium is not itself an MCP server. Agents use it through code, and community MCP wrappers exist outside the project.
Sources and freshness
- Selenium WebDriver documentation: https://www.selenium.dev/documentation/webdriver/
- SeleniumHQ/selenium repository: https://github.com/SeleniumHQ/selenium
Last checked 2026-10-07. Verification label: source verified, which means public claims trace to the sources above. It does not mean AgentsUse ran the tool. Spotted an error? Send a correction. Back to the tools directory.