The Browser Automation Guide: Playwright, Selenium, and CAPTCHA-Proofing Your Flows

Browser automation lets you control a real web browser with code. It is how teams test applications, scrape data, and run workflows that require full JavaScript execution. This guide is the hub for everything on the blog about browser automation, and it is where CAPTCHAs almost always enter the picture.

Why Browsers Instead of Plain HTTP?

Not every flow can be automated with requests or curl. You need a browser when:

• The site renders critical content with client-side JavaScript.
• Login or checkout flows depend on browser APIs.
• The site detects non-browser clients by TLS fingerprint or HTTP headers.
• CAPTCHA providers like reCAPTCHA require a real rendering environment.

A headless or headed browser executes the full page, just like a human visiting the site.

The Two Main Tools

Playwright

Playwright (Microsoft) is the modern choice. Highlights:

• Cross-browser: Chromium, Firefox, and WebKit with one API.
• Auto-waiting: locators wait for elements automatically, reducing flaky tests.
• Robust selectors: text, CSS, and role-based locators.
• Great debugging: traces, screenshots, and video recording.

from playwright.sync_api import sync_playwright

with sync_playwright() as p:
    browser = p.chromium.launch(headless=False)
    page = browser.new_page()
    page.goto("https://example.com")
    page.fill("#email", "[email protected]")
    page.click("button[type=submit]")
    browser.close()

Selenium

Selenium is the battle-tested veteran. It supports Java, Python, C#, and Ruby, and integrates with huge ecosystems (Grid, BrowserStack, Sauce Labs). Use it when your team or stack already depends on Selenium WebDriver.

from selenium import webdriver
from selenium.webdriver.common.by import By

driver = webdriver.Chrome()
driver.get("https://example.com")
driver.find_element(By.ID, "email").send_keys("[email protected]")
driver.find_element(By.CSS_SELECTOR, "button[type=submit]").click()
driver.quit()

Both are excellent. Pick Playwright for new greenfield automation and Selenium for legacy or cross-language needs.

Where CAPTCHAs Appear in Browser Flows

CAPTCHAs interrupt automation at the worst possible moment. Common spots:

• Login — before or after credentials are submitted.
• Signup — during account creation.
• Checkout — payment or order review steps.
• Search and filters — after repeated requests.
• Form submission — comment, contact, and upload forms.

When a CAPTCHA appears, your script must stop, solve it, and resume without losing state.

Handling CAPTCHAs in a Browser Flow

The robust pattern is:

1. Detect the challenge — poll for a selector unique to the provider (e.g., .g-recaptcha, .h-captcha, iframe[src*="recaptcha"]).
2. Extract the site key — read it from the DOM or network requests.
3. Solve externally — call the DeathByCaptcha API with the site key and challenge details.
4. Inject the token — use JavaScript to set the g-recaptcha-response textarea and fire the callback, or click the checkbox in a headed browser.
5. Resume — let the page continue and assert that the next step loaded.

Playwright example for reCAPTCHA v2:

# Assume `token` comes from the DeathByCaptcha API
page.evaluate(
    "document.getElementById('g-recaptcha-response').innerHTML = arguments[0];"
    "onSuccess(arguments[0]);", token
)

Making Your Automation Look Human

Sites evaluate behavior, not just puzzle answers. Reduce detection risk:

• Use persistent profiles — keep cookies, localStorage, and user-agent consistent.
• Human-like delays — randomize waits between actions.
• Real viewport and resolution — match a common device.
• Headless is fine, but headed is safer for the most protected sites.
• Use residential proxies — rotating datacenter IPs can trigger risk scoring.
• Avoid fingerprint leaks — disable automation flags or use stealth patches.

Scaling Browser Automation

As your flows grow:

• Run in parallel — one browser per worker, bounded by CPU and memory.
• Containerize — each worker in a container for isolation and cleanup.
• Centralize solving — one DeathByCaptcha API account for all workers, with per-task tracking.
• Instrument everything — log success rates, solve times, and failure reasons per site.
• Retry with exponential backoff — transient failures are normal.

Browser Automation and CAPTCHAs: The Bottom Line

A browser is a tool, not a silver bullet. The sites you automate will keep improving their bot detection. The winning combination is a modern browser driver, clean session hygiene, residential proxies, and a fast, accurate CAPTCHA solving service.

Next Steps

• CAPTCHA Guide — understand each challenge type before you automate it.
• API Learning Center — integrate solving directly into your scripts without a browser.
• Web Scraping Guide — build scrapers that survive anti-bot measures.

This guide is the hub for browser automation content. Check back for Playwright, Selenium, and framework-specific deep dives.