Extracting reCAPTCHA Parameters from Page Source

Every reCAPTCHA solve through CaptchaAI requires the correct sitekey and page URL. Some CAPTCHA types also need additional parameters — action (v3), data-s (Google sites), or enterprise flag. This guide covers every extraction method.

Parameters by reCAPTCHA version

Method 1: HTML attribute extraction

From data-sitekey attribute

import re
import requests

url = "https://example.com/login"
html = requests.get(url).text

# Find data-sitekey
match = re.search(r'data-sitekey=["\']([A-Za-z0-9_-]+)["\']', html)
if match:
    sitekey = match.group(1)
    print(f"Sitekey: {sitekey}")

# Check if invisible
invisible_match = re.search(r'data-size=["\']invisible["\']', html)
is_invisible = bool(invisible_match)
print(f"Invisible: {is_invisible}")

# Find callback
callback_match = re.search(r'data-callback=["\'](\w+)["\']', html)
callback = callback_match.group(1) if callback_match else None
print(f"Callback: {callback}")

# Check for data-s (Google-owned sites)
data_s_match = re.search(r'data-s=["\']([^"\']+)["\']', html)
data_s = data_s_match.group(1) if data_s_match else None
print(f"data-s: {data_s}")

JavaScript (Puppeteer)

const puppeteer = require('puppeteer');

const browser = await puppeteer.launch();
const page = await browser.newPage();
await page.goto('https://example.com/login', { waitUntil: 'networkidle2' });

const params = await page.evaluate(() => {
  const widget = document.querySelector('.g-recaptcha');
  if (!widget) return null;

  return {
    sitekey: widget.getAttribute('data-sitekey'),
    size: widget.getAttribute('data-size'),
    callback: widget.getAttribute('data-callback'),
    dataS: widget.getAttribute('data-s'),
    invisible: widget.getAttribute('data-size') === 'invisible',
  };
});

console.log(params);

Method 2: Script tag extraction

reCAPTCHA v3 and Enterprise sitekeys

v3 sitekeys are embedded in the script URL:

# Find sitekey from script src
v3_match = re.search(
    r'recaptcha/(?:api|enterprise)\.js\?.*?render=([A-Za-z0-9_-]+)',
    html
)
if v3_match:
    sitekey = v3_match.group(1)
    print(f"v3 Sitekey: {sitekey}")

# Check enterprise
is_enterprise = 'enterprise.js' in html
print(f"Enterprise: {is_enterprise}")

Finding the action parameter

The action is passed in JavaScript code, not in HTML attributes:

# Search for grecaptcha.execute calls
action_match = re.search(
    r'grecaptcha\.execute\s*\([^,]+,\s*\{[^}]*action\s*:\s*["\']([^"\']+)',
    html
)
if action_match:
    action = action_match.group(1)
    print(f"Action: {action}")

Method 3: Iframe src extraction

When reCAPTCHA is rendered inside an iframe:

# Find reCAPTCHA iframe
iframe_match = re.search(
    r'<iframe[^>]+src=["\']([^"\']*recaptcha/api2/anchor[^"\']*)["\']',
    html
)
if iframe_match:
    iframe_src = iframe_match.group(1)
    sitekey_match = re.search(r'k=([A-Za-z0-9_-]+)', iframe_src)
    if sitekey_match:
        sitekey = sitekey_match.group(1)
        print(f"Iframe sitekey: {sitekey}")

Method 4: JavaScript rendering extraction

For pages that render reCAPTCHA dynamically with grecaptcha.render():

# Find grecaptcha.render calls
render_match = re.search(
    r'grecaptcha\.render\s*\([^,]*,\s*\{([^}]+)\}',
    html
)
if render_match:
    config = render_match.group(1)
    sk = re.search(r'sitekey\s*:\s*["\']([A-Za-z0-9_-]+)', config)
    cb = re.search(r'callback\s*:\s*["\']?(\w+)', config)
    sz = re.search(r'size\s*:\s*["\'](\w+)', config)
    print(f"Sitekey: {sk.group(1) if sk else 'not found'}")
    print(f"Callback: {cb.group(1) if cb else 'not found'}")
    print(f"Size: {sz.group(1) if sz else 'not found'}")

Complete extraction function

import re
import requests

def extract_recaptcha_params(url):
    html = requests.get(url, timeout=15).text
    params = {"pageurl": url}

    # Sitekey from data-sitekey
    sk = re.search(r'data-sitekey=["\']([A-Za-z0-9_-]+)', html)
    if sk:
        params["sitekey"] = sk.group(1)

    # Sitekey from script render parameter (v3)
    if "sitekey" not in params:
        v3 = re.search(r'render=([A-Za-z0-9_-]{20,})', html)
        if v3:
            params["sitekey"] = v3.group(1)

    # Sitekey from iframe
    if "sitekey" not in params:
        iframe = re.search(r'recaptcha.*?k=([A-Za-z0-9_-]+)', html)
        if iframe:
            params["sitekey"] = iframe.group(1)

    # Sitekey from grecaptcha.render
    if "sitekey" not in params:
        render = re.search(r'sitekey\s*:\s*["\']([A-Za-z0-9_-]+)', html)
        if render:
            params["sitekey"] = render.group(1)

    # Version detection
    if re.search(r'data-size=["\']invisible', html):
        params["invisible"] = True
    if 'enterprise.js' in html:
        params["enterprise"] = True

    # Action (v3)
    action = re.search(
        r'action\s*:\s*["\']([^"\']+)',
        html[html.find('grecaptcha.execute'):] if 'grecaptcha.execute' in html else ''
    )
    if action:
        params["action"] = action.group(1)

    # data-s
    ds = re.search(r'data-s=["\']([^"\']+)', html)
    if ds:
        params["data_s"] = ds.group(1)

    # Callback
    cb = re.search(r'data-callback=["\'](\w+)', html)
    if cb:
        params["callback"] = cb.group(1)

    return params

# Usage
params = extract_recaptcha_params("https://example.com/login")
for k, v in params.items():
    print(f"  {k}: {v}")

Expected output:

  pageurl: https://example.com/login
  sitekey: 6Le-SITEKEY-abc123
  invisible: True
  callback: onCaptchaComplete

Submitting extracted parameters to CaptchaAI

data = {
    "key": API_KEY,
    "method": "userrecaptcha",
    "googlekey": params["sitekey"],
    "pageurl": params["pageurl"],
    "json": "1",
}

if params.get("invisible"):
    data["invisible"] = "1"
if params.get("enterprise"):
    data["enterprise"] = "1"
if params.get("action"):
    data["action"] = params["action"]
if params.get("data_s"):
    data["data-s"] = params["data_s"]

resp = requests.post("https://ocr.captchaai.com/in.php", data=data).json()

Troubleshooting

FAQ

Can I extract parameters without loading the page in a browser?

Yes, for most sites — the sitekey is in the HTML source. But JavaScript-rendered CAPTCHAs require a browser or headless browser.

Is the sitekey the same as the API key?

No. The sitekey is a public key assigned to the website. It's safe to share and is visible in the page source.

Use extracted parameters with CaptchaAI for reliable solving

Get your API key at captchaai.com.

Related guides

• Browser Console CAPTCHA Detection
• reCAPTCHA v2 Callback Mechanism
• reCAPTCHA v2 Invisible: Trigger Detection