Step 1 of 5
It starts with 700+ public lists.
Every hour a GitHub Action reads more than 700 public proxy lists. Most of what's in them is dead, duplicated or years old.
Every proxy on this list passed a real handshake, a honeypot check and a content check in the last run. The whole list is checked again every 6 hours.
Measured from GitHub's servers. From your own network some proxies will be faster, some won't answer at all.
working proxies right now. Out of about a million candidates, only these answered properly and passed every check.
tunnel HTTPS. They carry encrypted traffic through CONNECT, tested with a verified TLS handshake.
are elite. The target site can't tell that a proxy is in between: no forwarded IP, no Via header.
median response time. Half of the list answers faster than that. Sort by latency to get the quickest ones first.
Filter, sort, then copy or download exactly the ones you need. Click any address to copy it.
| Loading the list | |||||||
Step 1 of 5
Every hour a GitHub Action reads more than 700 public proxy lists. Most of what's in them is dead, duplicated or years old.
Step 2 of 5
Each candidate has to complete a proper HTTP, SOCKS4 or SOCKS5 handshake and fetch a page. An open port alone proves nothing.
Step 3 of 5
The proxy has to reach two unrelated sites and show an exit IP that isn't ours on both. Honeypots that fake a response drop out here.
Step 4 of 5
A static page has to arrive byte for byte. When we measured it, 54 of 270 working proxies slipped a script into the page. None of those make the list.
Step 5 of 5
HTTPS with verified TLS, anonymity, country, provider and whether the exit IP is on a spam blocklist are checked for everything that's left.
That's the list above. In an hour it starts over, and anything that stopped working is gone.
Free proxies come and go quickly. This is how the list changed over the last runs and where the proxies are right now.
One point every hour.
Where the exits are.
How fast they answered.
Run the same checks on your own machine, start from this list, or just fetch the plain text files.
700+ sources, all the checks, filters and a live dashboard in your terminal.
pipx install git+https://github.com/maximilianfeix/proxy-scraper.git proxy-scraper --want 50 --https-only
Starts from this list and checks it again from where you are. Takes about 30 seconds, then serves a rotating proxy on port 8899.
proxy-scraper --recheck live --serve
One proxy per line, fastest first, updated with every run.
The same files this page uses, next to it on GitHub Pages. Updated with every run, served with CORS, no key needed.
curl -s https://maximilianfeix.github.io/proxy-scraper/proxies.json | jq '.[0]'
{
"url": "socks5://203.0.113.10:1080",
"proxy": "203.0.113.10:1080",
"ptype": "socks5",
"latency": 92,
"country": "DE",
"https": true,
"anonymity": "elite",
"exit_ip": "203.0.113.10",
"asn": 14061,
"org": "DigitalOcean, LLC",
"hosting": true,
"blocklisted": false,
"streak": 5
}