# Proxies

Scrape.it does not provide proxies. Every scrape uses exactly one caller-owned source: a verified saved `proxyConnectorId`, or an authenticated request-scoped `proxy` URL for synchronous work.

Launch providers are Bright Data, Oxylabs, Decodo, SOAX, IPRoyal, Webshare, DataImpulse, generic authenticated proxies, and DIY mobile proxies. Username/password proxy authentication is supported; IP-whitelist-only accounts are not supported in the first privacy-aware release. SOCKS users should use `socks5h://` so target DNS is resolved through the proxy. Vendor host/port/username encodings to paste at signup are on https://scrape.it/connectors.md.

Gateways are resolved safely, private/reserved endpoints are rejected, and verification authenticates through a request-scoped broker lease plus signed neutral IP-address probes. Static and matching pre/post session probes can support exact attribution. Per-request rotation, mismatched probes, and unverified samples update provider/product/pool intelligence without pretending that the sampled IP handled a target request. Ordinary jobs never switch connectors without explicit authorization. Cross-user proxy pooling is forbidden.