Historical baseline

21-site scraping stress test

These runs happened on 24–25 May 2026 from one network, with no proxy. They show what each exact mode returned then. They do not tell you which tool will work best today.

12 run modes21 target URLs252 attemptsnone proxy

An attempt passed its recorded status and text-marker checks. HTTP 200 alone was not enough. Some targets needed JavaScript that HTTP clients cannot run; this was a stress test, not a fair all-purpose leaderboard.

Download published attempt JSONRead the method

Choose a run to inspect

Tool and modeVersionDateObserved passesConditions
requests2.34.22026-05-242/21 checksOne network · no proxy
httpx0.28.12026-05-243/21 checksOne network · no proxy
wreq0.11.32026-05-242/21 checksOne network · no proxy
curl cffi0.15.02026-05-245/21 checksOne network · no proxy
scrapling0.4.82026-05-245/21 checksOne network · no proxy
playwright1.60.02026-05-243/21 checksOne network · no proxy
patchright1.60.02026-05-243/21 checksOne network · no proxy
seleniumbase4.49.22026-05-244/21 checksOne network · no proxy
camoufox0.4.112026-05-244/21 checksOne network · no proxy
camoufox tuned0.6.02026-05-2511/21 checksOne network · no proxy
camoufox headed0.4.112026-05-243/21 checksOne network · no proxy
botosaurus4.0.972026-05-246/21 checksOne network · no proxy

Only the reviewed attempt metadata is published. Old raw page bodies and screenshots remain in the local lab and are not linked here.