🕐 Free Proxy Lists on a 5-Minute Refresh — Plus the 21 Self-Hosted Tools That Run the Loop on Your Own Clock

Tools & Scripts
, ,
1

Roadside stall selling bottles stamped checked and working, all filled from one tank, while a woman taps the spring herself
Roadside stall selling bottles stamped checked and working, all filled from one tank, while a woman taps the spring herself1672×941 667 KB

:one_o_clock: Every list refreshes on somebody else’s clock — these 21 tools run the loop on yours, free, no account

5 min :black_small_square: 15 min :black_small_square: 20 min :black_small_square: 30 min :black_small_square: 3 hrs :black_small_square: 6 hrs :black_small_square: yours

A proxy list is a photograph. What you actually want is the camera. Every tool below is sorted by the one spec nobody puts on the box: how often it goes and gets you new ones — starting with lists you just open in a browser, ending with a pool on your own machine that never stops.

Sixty seconds, nothing to sign up for:

pip install "fast-proxy-checker[socks]"
proxy-checker fetch --count 25 --type socks5 --output alive.txt

That pulls from public sources and checks them in one go — you end up with a file of proxies plus their speed and their anonymity level. Two words to know before you go any further:

elite — the site sees the proxy and nothing else. This is what you want.
transparent — the site sees the proxy and your real IP. It “works” and protects nothing.

:1st_place_medal: GRAB — no install, it’s just a URL

Open the link, copy the file, done. Sorted fastest first.

ClearProxy — every 5 minutes, checked HTTP / SOCKS4 / SOCKS5 with geo, anonymity and ISP attached. The clever bit: it also re-tests everything against six real sites, so there are separate files for Discord, Google, Instagram, Reddit, TikTok and X. Whether a proxy reaches your target is the only test that ever counted, and it’s already done.

ProxRipper — every 15 minutes, all four protocols, four plain text files:

full_proxies/http.txt   ·   https.txt   ·   socks4.txt   ·   socks5.txt

CX2F/proxy-scraper — every 3 hours, published as TXT, CSV and JSON so you can drop it straight into a spreadsheet or a script.

nestorchurin/proxy-list — every 6 hours, checked and deduplicated, split by protocol and by source.

BLITZPROXY — every 2 days as a file, but it’s also the whole machine: docker compose up -d gives you a web GUI on port 3000 where you start and stop the checker yourself, and ProxyURL.json is where you paste your own sources.

Then take the clock off them

Five minutes is fast. It’s still somebody else’s five minutes, aimed at somebody else’s sources. Everything below runs on yours.

🌾 HARVEST — go and get the candidates · 6 tools

ProxyGather — 126★ and the big one. python ProxyGather.py run scrapes and checks in a single command, its Action regenerates lists every 30 minutes, --only and --exclude pick which sources fire, and the README has an Adding Your Own Sites section.

proxine — 86 sources across 220 endpoints in one shot, and the only one that tracks how fresh each source is: it reads each source’s real last-update time and -F 86400 drops any that hasn’t moved in a day. Its author is straight with you — it aggregates, it doesn’t grade, so pipe it into a checker.

Raven Proxy Scraper — MIT, dozens of sources including GitHub lists and live APIs, multi-threaded validation, and it writes you a ready proxychains.conf plus a cron example.

ProxyHawk — skips lists altogether if you’d rather: -discover -discover-source shodan|censys|freelists|webscraper, with honeypot filtering and -discover-countries US,GB,DE. It also runs SSRF and 55+ CVE checks, so it tells you when a “proxy” is bait.

Hiddence/ProxyScraperChecker — Go, Docker, live progress bar, strict mode, and it folds whatever is already in your /out folder back into the next run instead of starting from zero.

IlmLV/proxy-scraper — a PHP library rather than an app: 36+ built-in sources, per-source cron schedules, checks anonymity, latency, header leaks and methods. Drop it inside something you already run.

🔎 JUDGE — kill the corpses, catch the leakers · 5 tools

fast-proxy-checker — the pip one at the top. 500+ a minute, labels elite / anonymous / transparent plus country, works as a CLI or as an import.

Proxy Reaper — speed buckets from ultrafast to slow, GeoIP and optional reverse DNS, exports to JSON, CSV and SQLite, autosaves so a crash doesn’t cost you the run. The one to put on a schedule.

ProxyAtlas — two Go binaries and the sharpest idea in the kit: it scores your sources, not just your proxies. A source that keeps handing back junk earns a fail-threshold and a 12-hour cooldown, so over a few weeks your source list quietly stops looking like everyone else’s. Adaptive fallback for the nights strict mode returns nothing.

proxy-finder — zero-setup CLI, 8 output formats, plugs into what you already use:

python3 skill/proxy_finder.py -n 5 -t socks5 -f proxychains

:adhesive_bandage: On Ubuntu / Debian 24.04 its auto-installer hits externally-managed-environment. Run it from a virtualenv and it’s fine.

Proxy-Hunter — a Python library, not a script: caching, pool warming, rotation and concurrent acquisition, so your own code asks for a proxy instead of reading a file.

🔌 SERVE — one address that's always alive · 5 tools

Rota — 566★, Apache-2.0, the literal answer to “new proxies every X minutes”: you add source URLs each with its own refresh interval in minutes, a background scheduler re-fetches whatever’s overdue, unhealthy nodes drop out by themselves, and requests rotate random / round-robin / least-connections / time-based. Dashboard, metrics, chains through Burp or ZAP.

git clone https://github.com/alpkeskin/rota.git && cd rota
docker compose up -d
make password        # the admin password is generated into the logs

Magpie — from the author of KC-Checker. Every proxy carries a reputation score built from uptime, latency and anonymity, and you spin your own rotating endpoints out of the healthy pool. Multi-user, one-line installer:

curl -fsSL https://raw.githubusercontent.com/Kuucheen/magpie/refs/heads/master/scripts/install.sh | bash

ProxyPoolWithUI — 308★, 14+ sources, validates in real time, browser UI and an API. The turnkey pick if you want a pool without reading a config file.

Proxy-HTTP-SOCKS-Get-Pool — MIT, C#, for the Windows crowd: built-in scrapers, speed, anonymity and geo tests, filtering, sorting and auto-rotation in one desktop tool.

GhostTrace — multi-hop chains, identity rotation, FastAPI backend with a live dashboard. Born June 2026, so sandbox it before you trust it with anything.

The part nobody does: make the source set yours

Everyone points the same scrapers at the same hundred lists, so everyone ends up dialling the same tired IPs. Five ways out, all in the kit above: BLITZPROXY’s ProxyURL.json, ProxyGather’s own-sites section, Rota’s per-source URL and interval, proxine’s -F flag binning any source that hasn’t moved in a day, and ProxyAtlas quietly scoring which of your sources still pay out. ProxyHawk goes furthest and pulls candidates straight out of Shodan and Censys instead of reading anyone’s list at all.

:warning: A public proxy is a stranger’s machine. Nothing with a login, a card, a token or a session cookie goes over one — and elite only describes what the proxy hides, never who owns it.

:toolbox: Companion piece — what “working” actually means (it’s six tests, not one) and the full checker toolbox: Stop Downloading Dead Proxies — Build the Filter That Only Keeps Working Ones

Let the list rot. Yours is sixty seconds old.