Proxy tier benchmark Protection systems

Why does a scraper get HTTP 202 with an empty body?

HTTP 202 Accepted, content-length zero. Nothing is refused, nothing is challenged, and nothing is served, in about 130 milliseconds.

This is the most dangerous outcome in the dataset for anyone building a pipeline, because it is a success by every conventional measure. A scraper that branches on response.ok records a clean run and writes empty rows. A monitoring dashboard built on status codes shows 100% availability while the collection produces nothing. There is no exception to catch and no error page to notice.

Portals behind HTTP 202 with an empty body in this census

Sixty-one candidate URLs were probed once each from a clean US datacenter IP on 2026-08-18; 2 of the portals below were then measured through all six proxy arms, ten rounds each.

Portal Measured in Behaviour
us_in_bsd_search six-arm run HTTP 202, zero bytes, on all 60 requests — identical unproxied, datacenter, anti-detect and residential; median 129 ms
us_nc_realestate six-arm run HTTP 202, zero bytes, on 50 of 60 requests; the residential arm got HTTP 404 with a 34 KB reCAPTCHA page instead

Both are the same kind of endpoint

Indiana's business search and North Carolina's real estate licensee search are both ASP.NET search applications behind an edge layer that answers a non-browser request by accepting it and dropping it. The reconnaissance probe found both the same way, weeks of scraper debugging apart, and the harness classifies an empty 2xx body as a soft block before it evaluates the content assertion for exactly this reason.

North Carolina picks its defence by exit type

The interesting half of North Carolina's result is what the residential arm got: HTTP 404 with a 34,196-byte reCAPTCHA page, on the same URL, in the same three-hour window, while every unproxied, datacenter and anti-detect request got the empty 202. The site chose a different defence for a different network.

That is corroborated from the compliance gate, which recorded ncrec.gov serving its robots.txt normally to a US datacenter IP and returning 403 to a European residential IP.

Practical note

Assert on content, never on status. A benchmark or a monitor that scores HTTP status codes will report these two targets as fully available at every proxy tier, and will be wrong on all 110 requests.


Run date: 2026-08-18 · All protection systems · All 32 benchmarked targets · Full report · Census data


Test your own target

The result above describes one URL on one day. Paste your own URL into the live tester and see the real response before creating an account.

Test your target See plans