Skip to content
CrawlfolioSourced research on web data for AI

Evidence

Evidence runs

What an evidence run is, and why none are published on Crawlfolio yet. Every "I ran" claim here will link to one.

An evidence run is a small, scripted check that I log in full. A fixed script sends a capped number of requests to one vendor, on free or trial credits, and points only at test pages I host with known correct answers or at sandboxes built for scraping practice. The script version, settings, raw results and a short summary are kept for every run.

Each one is a small-sample check, not a benchmark. A few dozen requests against test pages can't predict what you'll see at your volume or on the sites you care about.

No runs are published yet. Everything on this site so far is research: I read pricing pages, docs and community threads, and link each claim to its source. When a run is published it will be listed here, and every "I ran…" claim on the site will link to it.

The rules every run follows, from terms checks to fixed settings, are on the methodology page.