Web Unblocker

A web unblocker is a product category, not a technique: one endpoint that bundles proxy selection and rotation, TLS and header fingerprint management, headless rendering, challenge and CAPTCHA handling, and retry logic behind a single request. You send a URL and get HTML back. Billing is success-based, though the unit varies by vendor. Bright Data prices Web Unlocker at $1.5 per 1,000 requests pay-as-you-go, with a $499 Scale plan including 383K requests and $1.3 per 1,000 after that, and states "pay only for success" (brightdata.com/pricing/web-unlocker, read 2026-07-27). Oxylabs bills Web Unblocker per gigabyte instead, at $5.64/GB on its Micro plan down to $4.50/GB on Advanced, also on a success-only basis (oxylabs.io/products/web-unblocker, read 2026-07-27). Scrapfly uses a credit balance where the credit cost of a request rises with JavaScript rendering, residential proxies and country targeting, and exposes a cost estimator plus a cost=true API parameter so you can price a request before running it (scrapfly.io/pricing, read 2026-07-27).

What the abstraction hides is fingerprint parity against a defense that moves. Keeping a TLS handshake, HTTP/2 frame ordering, header order, and browser API surface mutually consistent is ongoing work, and it degrades quietly: a stack that worked last quarter starts returning challenge pages this quarter with no code change on your side. Paying someone whose full-time job that is, is the whole value proposition, and it is a reasonable trade.

The costs are real. Failures are opaque, so when a target stops working you cannot tell whether the proxy, the fingerprint or the renderer failed, and your only lever is a support ticket. Per-request pricing runs several times the raw bandwidth cost of the same fetch. Request formats are vendor-specific, so switching providers means rewriting the call sites. And a vendor that quotes a headline success rate is quoting it against its own target mix, not yours.

The math flips on your own success rate. At Bright Data's $1.5 per 1,000, one request costs $0.0015. The same 100 KB page pulled over Decodo residential at $3.25/GB on its 25 GB tier (decodo.com/pricing, read 2026-07-27) costs about $0.0003 in bandwidth, roughly five times less, but only if it succeeds first try. Raw proxies bill retries and failures too, so at a 40% success rate that $0.0003 becomes about $0.0008, and the remaining gap has to cover your engineering time. Run unblockers on the handful of targets that genuinely fight back, run raw proxies plus your own retry stack on everything else, and recheck the split when either number moves.