How to use a proxy with wget

Setup guide · Last revised Oct 20, 2026

The quickest way

https_proxy="http://USERNAME:[email protected]:8000" wget -q -O - https://api.ipify.org

prints the address the site saw:

203.0.113.24

Run it again without the https_proxy=... part: if the address is the same, wget is not using the proxy. https_proxy is for https:// pages and http_proxy for http:// pages; wget’s manual says they “should contain the URLs of the proxies for HTTP and HTTPS connections respectively” (GNU Wget manual, Proxies). Set both if your job fetches both.

The login on its own

A password with @, : or # breaks the address. Pass the login separately:

wget -q -O - -e use_proxy=yes -e https_proxy=proxy.example.com:8000 \
  --proxy-user=USERNAME --proxy-password=PASSWORD https://api.ipify.org

wget sends it with Basic authentication; the manual notes that “For proxy authorization only the Basic authentication scheme is currently implemented”, which is what proxy providers use. To keep the login out of commands, put https_proxy, proxy_user and proxy_password in your ~/.wgetrc, readable only by you.

Turn it off for one run

--no-proxy makes wget connect directly “even if the appropriate *_proxy environment variable is defined” (manual). no_proxy=".example.org" skips the proxy for one domain and everything under it.

What wget prints when it fails

Downloading many files

wget’s recursive mode (-r) can fetch a whole site through the proxy, and each file counts against a plan billed by the GB. Limit it to what you need (-l 1 for one level, -A pdf for one file type, --wait=1 between requests) before you start. What a large crawl costs: how much bandwidth web scraping uses.

Next

How to test a proxy checks the location and speed. For the codes a site can send back: proxy errors.