Skip to content

captchas.md

Latest commit

 

History

History
98 lines (78 loc) · 3.17 KB

captchas.md

File metadata and controls

98 lines (78 loc) · 3.17 KB
Raw

Captcha auto-solve

Pass a funsolver.com API key to funbrowser.start and detected captchas solve themselves on every tab. No glue code required.

async with await funbrowser.start(
    api_key="fs_xxx",
    auto_solve=True,   # default
) as browser:
    tab = await browser.get("https://protected-with-captcha.com")
    # if the page has a Turnstile / reCAPTCHA / hCaptcha widget, the
    # solver picks it up automatically and writes the token before
    # your next action.
    await tab.click("button[type=submit]")

Supported types

Captcha Detector matches Response field written
Cloudflare Turnstile .cf-turnstile[data-sitekey] textarea[name="cf-turnstile-response"]
reCAPTCHA v2 (+ Enterprise) .g-recaptcha[data-sitekey] every g-recaptcha-response* textarea
reCAPTCHA v3 (+ Enterprise) hooks window.grecaptcha.execute() returned to page-script .then()
hCaptcha .h-captcha[data-sitekey] textarea[name="h-captcha-response"]
FunCaptcha / Arkose [data-pkey] input[name="fc-token"] and aliases
GeeTest v3 + v4 hooks window.initGeetest / initGeetest4 geetest_challenge / _validate / _seccode form fields, success callback invoked with the validated payload

Manual solve API

If you want to drive the solver from script code instead of relying on the page-side detector:

client = funbrowser.FunSolverClient("fs_xxx")
token = await client.solve_recaptcha_v2(
    sitekey="6Lc...",
    page_url="https://example.com",
    invisible=False,
)
# then write `token` into the textarea / trigger the callback yourself.

All methods are async. Each maps to its 2captcha / anti-captcha task type (RecaptchaV2TaskProxyless, HCaptchaTaskProxyless, etc.). The client polls until the task is ready and returns the token string.

Custom integration

If the page's widget doesn't match the default detector selectors, you can either:

  1. Add a data attribute to the widget so our scanner finds it (the detectors use the documented attributes — data-sitekey, data-pkey, etc. — which most pages already have)

  2. Call __funbrowser.solve() from your own page JS:

    const token = await window.__funbrowser.solve({
  type: "recaptcha2",
  sitekey: "6Lc...",
  url: location.href,
  invisible: false,
});

  3. Solve from Python and inject yourself:

    tok = await browser.solver_client.solve_hcaptcha(
    sitekey="...",
    page_url=tab.url,
)
await tab.evaluate(f"""
    document.querySelector('[name="h-captcha-response"]').value = {tok!r};
""")

Activity log

The local web panel (pip install funbrowser[panel]) shows every captcha solve as it happens, with the captcha type, browser index, and time taken. The panel also displays your live FunSolver balance.

Endpoint configuration

If the real funsolver.com endpoints differ from the 2captcha-style defaults the client assumes, override:

await funbrowser.start(
    api_key="fs_xxx",
    solver_base_url="https://api.funsolver.com",  # default
)

Or subclass FunSolverClient and override _post / _create_task / _solve for full control.