Unblocking tiers

none (default, +0)

A datacenter IP and honest, standard browser defaults. Most public pages load fine this way, and it is the cheapest and fastest option.

stealth (+5)

A consistent real-Chrome fingerprint: user agent, platform, client hints, locale, timezone and WebGL vendor that agree with each other, with common automation leaks patched. It helps with sites that reject obvious headless browsers.

residential (+0, your own proxy)

The stealth identity, sent out through a residential proxy that your team brings: Decodo, Oxylabs, Bright Data, IPRoyal, SOAX, Evomi or any HTTP, HTTPS or SOCKS5 proxy. Save its URL in Settings → Proxy, then pick an exit country per request if you like. You pay your provider for the bandwidth and we add no surcharge, on every plan including the free one. See Bring your own proxy.

Without a team proxy, unblock: "residential" returns 403 proxy_not_configured. When the proxy itself fails (bad credentials, unreachable), you get 502 proxy_failed with a reason, and are not charged.

auto

Tries none, then stealth, then residential (when your team has a proxy), stopping at the first tier that is not blocked, or at the first CAPTCHA. The response's tier_used says which one worked and attempts lists each try. You are charged for the tier that succeeded only; failed attempts are free.

CAPTCHAs

We do not solve CAPTCHAs, click challenge widgets or use solving services, on any plan or tier. Some sites run a JavaScript check that any normal browser passes by loading the page and waiting (Cloudflare's "Just a moment", for example). With mode: "auto" or "browser", when a page answers with a check, we load it once more in our standard browser, on the same tier, and wait up to 12 seconds for the site's own check to finish, if robots.txt allows the URL. We don't change our identity or move to another tier to get past it.

If the check doesn't clear, or it asks for interaction (reCAPTCHA, hCaptcha, Turnstile, press-and-hold and similar), the request stops right there: no further tiers, no retries. You get 422 captcha_required, with provider naming the CAPTCHA or vendor when we can tell, and are not charged.

When everything fails

If the target blocks every tier we tried, you get 422 target_blocked with block_reason and block_vendor, and are not charged. We detect blocks from 403 and 429 responses, vendor challenge pages, pages that say you were blocked, and pages with no visible text at all. CAPTCHAs end the request earlier, as above, and so does a redirect to a login page (422 login_required): we never log in.