Selenium Captcha Solver API

Why Selenium Gets Blocked by CAPTCHAs

reCAPTCHA, hCaptcha, and Cloudflare Turnstile are specifically designed to detect browser automation. Even with undetected-chromedriver or selenium-stealth, high-traffic sites will eventually challenge your bot. Fighting the detection is an arms race you will lose — the professional approach is to solve the challenge programmatically and continue your automation flow.

Step 1 — Install the Dependencies

pip install selenium captchakings
# Optional but recommended for fewer detections:
pip install undetected-chromedriver

Step 2 — reCAPTCHA v2 Token Flow (Planned)

This flow goes live when reCAPTCHA support launches — see /recaptcha-solver. The image-captcha flow below works today. The preview example opens a page, extracts the sitekey, solves it with CaptchaKings, and injects the token:

import time
from selenium import webdriver
from captchakings import CaptchaKings

API_KEY  = "ck_your_api_key_here"          # from captchakings.com/dashboard
PAGE_URL = "https://example.com/signup"     # page with the reCAPTCHA

client = CaptchaKings(API_KEY)
driver = webdriver.Chrome()
driver.get(PAGE_URL)

# 1. Extract the sitekey rendered on the page
sitekey = driver.execute_script("""
    const el = document.querySelector('[data-sitekey]');
    return el ? el.getAttribute('data-sitekey') : null;
""")
print("sitekey:", sitekey)

# 2. Ask CaptchaKings to solve it (returns a valid g-recaptcha-response token)
token = client.solve_recaptcha_v2(sitekey=sitekey, page_url=PAGE_URL)
print("token received")

# 3. Inject the token and submit
driver.execute_script("""
    document.getElementById('g-recaptcha-response').innerHTML = arguments[0];
    document.getElementById('g-recaptcha-response').style.display = 'block';
""", token)
driver.execute_script("document.querySelector('form').submit()")

time.sleep(3)
print("Done:", driver.current_url)
driver.quit()

Step 3 — Solve Image / Text CAPTCHAs (3 Lines)

For classic image captchas, screenshot the element and send it straight to the API:

img = driver.find_element("css selector", "#captcha-image")
img.screenshot("captcha.png")

result = client.solve("captcha.png")
print(result["prediction"], "| confidence:", result["confidence"])

Supported CAPTCHA Types in Selenium Flows

Pro Tips for Production Selenium Bots

1. Reuse sessions. Solve the captcha once, then persist cookies (driver.get_cookies()) instead of re-solving on every run.
2. Rotate residential proxies. Pair your solver with the CaptchaKings IP Rotator so the IP that solves the captcha matches the IP that submits the form.
3. Handle failures idempotently. Failed solves are not billed — wrap the call in a retry with exponential backoff and you only pay for successes.
4. Inject, don't click. Token injection via execute_script is far more reliable than simulating clicks on the reCAPTCHA iframe.

Frequently Asked Questions

How do I bypass reCAPTCHA in Selenium Python?

reCAPTCHA support is planned, not yet live — see /supported-captchas and /recaptcha-solver. At launch you will extract the sitekey, send it to CaptchaKings, inject the returned token into #g-recaptcha-response, then submit the form — exactly as previewed in Step 2 above. Image/text captchas work today via Step 3.

Does this work with undetected-chromedriver and headless mode?

Yes. The solver runs server-side, so it works with any WebDriver flavor: headless Chrome, Firefox, undetected-chromedriver, and remote Selenium Grid / BrowserStack sessions.

What happens if the token is rejected?

Call client.report_bad(task_id) for an automatic refund of that solve, then request a fresh token. Incorrect solves can always be reported for a refund.

Related Integration Guides