Get started
Introduction
Scrapivore scrapes through rotating proxies with one API call. It picks proxies for you, races several in parallel, and returns the first good response. Proxies that fail on a site rest for a while, and the ones that work get picked more often. Everything is a plain HTTP API, so it works from any language.
2
Copy your token
Log in and it's filled into every example.
Guides
Scrape a URLOptions, response shape, retries.
Scrape 50 URLs at onceOne call, results in input order.
Keep the same IPSticky sessions for logins and carts.
Request templatesTurn a copied cURL into a GET URL.
Use your own proxiesPaste a list or connect Webshare.
Errors and limitsStatus codes, rate limits, retries.
How a request runs
- Pick. The engine takes the best-ranked proxies for the target domain from the pool you chose.
- Race. It sends
race_widthrequests at once (3 by default). - Return. The first 2xx comes back with the proxy used and its latency. The other attempts are cancelled.
- Learn. Proxies that failed go on a per-domain cooldown. If nothing wins, another round runs (up to 2).