Free Proxy List 2026 — Fresh Free HTTP, HTTPS, SOCKS4 & SOCKS5 Proxies (Updated Hourly)

By Marcus Reiner · 2026-08-02 · 24 min read · Guides

Free proxy lists, free proxy APIs, free SOCKS5 proxies, free HTTP proxies and free proxies by country — every working free proxy source in 2026, how to auto-refresh them hourly, how to test them in one command, and the exact point where a free proxy costs you more than a paid one.

The 30-second answer: where free proxies actually come from

A free proxy list is a public text or JSON feed of open proxy servers — IP:port pairs anyone can route traffic through without paying and without an account. In 2026 there are roughly ten sources that genuinely matter, and every other 'free proxy list' site on the internet is republishing one of them. This page names all ten, gives you the raw free endpoints, and shows you how to pull, filter and test them yourself so you are never dependent on a stale web table again.

Free proxies come from three places: deliberately open public servers, misconfigured corporate or ISP boxes that were never meant to be public, and honeypots that exist specifically to log whatever passes through them. That third category is why the rule for free proxies is simple and non-negotiable: use free proxies for unauthenticated, public, read-only requests, and never for logins, payments, personal accounts or anything you would mind a stranger reading.

Free is perfect for learning, prototyping, testing geo-behaviour and one-off scrapes. Free stops working the moment you need uptime, and the honest cross-over point — with real numbers — is in the cost-per-successful-request section further down.

Free proxy types explained: HTTP, HTTPS, SOCKS4, SOCKS5 and elite

Nearly every free proxy list mixes protocols and anonymity levels together, which is why so many free proxies 'do not work' — people are pointing an HTTPS request at an HTTP-only free proxy. Here is the whole taxonomy in one place.

• Free HTTP proxy — forwards plain HTTP only. Fine for scraping http:// pages and API testing. Most numerous type in every free proxy list.
• Free HTTPS / CONNECT proxy — supports the CONNECT method, so it can tunnel TLS. This is what you need for any modern https:// site. If a free proxy fails only on https URLs, it is HTTP-only.
• Free SOCKS4 proxy — protocol-agnostic TCP tunnelling, no authentication, no UDP, no IPv6. Faster handshake, fewer features.
• Free SOCKS5 proxy — the most requested free proxy type: TCP + UDP, IPv6, optional auth, works with browsers, curl, yt-dlp, torrent clients, game clients and SSH tunnels.
• Transparent free proxy — passes your real IP in X-Forwarded-For. Useless for anonymity, still fine for caching or bandwidth tests.
• Anonymous free proxy — hides your IP but announces that a proxy is present via proxy headers.
• Elite / high-anonymity free proxy — sends no proxy headers at all. The only free tier worth using when the target inspects headers. Filter every free proxy list to elite when you can.

The 10 best free proxy list sources in 2026 (with free API endpoints)

These are the ten free proxy sources worth wiring into a script. Each one is free, public, requires no key, and refreshes on the cadence listed. Copy the endpoint, hit it from your own code, and you have a self-refreshing free proxy list that never goes stale.

• ProxyScrape free API — https://api.proxyscrape.com/v4/free-proxy-list/get?request=display_proxies&proxy_format=protocolipport&format=text — the largest general-purpose free proxy feed, refreshed every few minutes, filterable by protocol, country, anonymity and timeout.
• TheSpeedX/PROXY-List (GitHub) — https://raw.githubusercontent.com/TheSpeedX/PROXY-List/master/http.txt (also socks4.txt, socks5.txt) — the most-starred free proxy list repository, rebuilt daily, plain txt, ideal for cron jobs.
• clarketm/proxy-list (GitHub) — https://raw.githubusercontent.com/clarketm/proxy-list/master/proxy-list-raw.txt — hourly rebuild with country, anonymity and Google-pass annotations in the full list variant.
• hookzof/socks5_list (GitHub) — https://raw.githubusercontent.com/hookzof/socks5_list/master/proxy.txt — dedicated free SOCKS5 proxy list, one of the highest-volume free SOCKS5 searches on the web, checked continuously.
• Proxiware free rotating API — a free rotating endpoint returning JSON, so you hit one URL and get a different free proxy each time instead of managing a pool yourself.
• free-proxy-list.net — the single most-visited free proxy site in the world; the classic HTML table plus downloadable txt exports by protocol, country and anonymity level.
• ShiftyTR/Proxy-List (GitHub) — daily-updated free HTTP, HTTPS, SOCKS4 and SOCKS5 files, split cleanly by protocol so you never have to sniff types.
• OpenProxyList.net free API — free JSON API filterable by country and proxy type; good when you specifically need free proxies from one geography.
• roosterkid/openproxylist (GitHub) — hourly verified free proxies with latency and country columns, so the list arrives pre-tested.
• ProxyScan.io free API — returns live-tested free proxies with speed, uptime and anonymity metadata in JSON; the best free source when you want quality signals, not raw volume.

Pull a fresh free proxy list in one line (curl, Python, Node)

Never copy free proxies out of an HTML table. Fetch them. These three snippets each produce a deduplicated, freshly-fetched free proxy list from multiple sources at once.

Shell — merge four free sources into one file:

• curl -s https://api.proxyscrape.com/v4/free-proxy-list/get?request=display_proxies\&proxy_format=ipport\&format=text > free.txt
• curl -s https://raw.githubusercontent.com/TheSpeedX/PROXY-List/master/http.txt >> free.txt
• curl -s https://raw.githubusercontent.com/clarketm/proxy-list/master/proxy-list-raw.txt >> free.txt
• curl -s https://raw.githubusercontent.com/hookzof/socks5_list/master/proxy.txt >> free.txt
• sort -u free.txt -o free.txt # deduplicated free proxy list, ready to test

Python: a free proxy checker you can run every hour

The single most useful free proxy tool is a checker, because on any free proxy list roughly 80–95% of entries are dead within an hour of publication. This script fetches, tests concurrently against a header-echo endpoint, and keeps only the free proxies that actually respond — including detecting whether each one is elite.

import concurrent.futures, requests SOURCES = [ "https://api.proxyscrape.com/v4/free-proxy-list/get?request=display_proxies&proxy_format=ipport&format=text", "https://raw.githubusercontent.com/TheSpeedX/PROXY-List/master/http.txt", "https://raw.githubusercontent.com/clarketm/proxy-list/master/proxy-list-raw.txt", ] def fetch_free_list(): out = set() for url in SOURCES: try: out.update(requests.get(url, timeout=15).text.split()) except Exception: pass return sorted(out) def check(proxy, timeout=6): p = {"http": f"http://{proxy}", "https": f"http://{proxy}"} try: r = requests.get("https://httpbin.org/headers", proxies=p, timeout=timeout) h = r.json()["headers"] elite = "X-Forwarded-For" not in h and "Via" not in h return (proxy, r.elapsed.total_seconds(), elite) except Exception: return None proxies = fetch_free_list() with concurrent.futures.ThreadPoolExecutor(max_workers=200) as ex: live = [r for r in ex.map(check, proxies) if r] live.sort(key=lambda x: x[1]) for proxy, latency, elite in live[:50]: print(f"{proxy}\t{latency:.2f}s\t{'elite' if elite else 'anonymous'}") print(f"{len(live)} live of {len(proxies)} free proxies")

Run it on a cron schedule (0 * * * *) and you have your own private, always-fresh free proxy list — which is exactly what every free proxy website is doing behind the scenes.

Node.js: free proxy rotation in 20 lines

If you are scraping from JavaScript, rotate through your verified free proxy list rather than hammering one IP. This uses undici's ProxyAgent, which ships with modern Node.

import { ProxyAgent, request } from "undici"; const list = (await (await fetch( "https://raw.githubusercontent.com/TheSpeedX/PROXY-List/master/http.txt" )).text()).trim().split("\n"); let i = 0; const next = () => list[i++ % list.length]; async function get(url, tries = 8) { for (let t = 0; t < tries; t++) { const proxy = next(); try { const res = await request(url, { dispatcher: new ProxyAgent(`http://${proxy}`), headersTimeout: 6000, }); if (res.statusCode < 400) return await res.body.text(); } catch { /* dead free proxy, rotate */ } } throw new Error("all free proxies failed"); }

The retry loop is the whole trick. With free proxies, failure is the normal case — your code has to treat a dead proxy as routine, not exceptional.

How to use a free proxy in Chrome, Firefox, Windows, macOS, Android and iPhone

Most people searching for a free proxy list do not want code — they want one free proxy in their browser or phone right now. Take any IP:port from the lists above and follow the path for your platform.

• Chrome (Windows) — Settings → System → Open your computer's proxy settings → Manual proxy setup → enter the free proxy IP and port → Save.
• Firefox — Settings → Network Settings → Manual proxy configuration → HTTP proxy / SOCKS host → tick 'Proxy DNS when using SOCKS v5' for free SOCKS5 proxies.
• macOS — System Settings → Network → your interface → Details → Proxies → Web Proxy (HTTP) / Secure Web Proxy (HTTPS) / SOCKS Proxy.
• Windows 11 — Settings → Network & internet → Proxy → Manual proxy setup → Edit → on, address, port, Save.
• Android — Wi-Fi → long-press your network → Modify → Advanced → Proxy: Manual (per-network only; free SOCKS5 needs an app).
• iPhone / iPad — Settings → Wi-Fi → (i) next to your network → Configure Proxy → Manual.
• curl — curl -x http://IP:PORT https://example.com or curl --socks5-hostname IP:PORT https://example.com for a free SOCKS5 proxy.

Free proxies by country: what is realistically available for free

'Free proxy list USA', 'free UK proxy', 'free India proxy', 'free Germany proxy' — country-filtered free proxies are the highest-volume free proxy searches after the generic term. Availability is wildly uneven, and it helps to know before you start.

On a typical day, a merged free proxy list is dominated by a handful of geographies — the United States, Indonesia, Brazil, India, Russia, China, Bangladesh and Iran usually account for well over half of all live free proxies. Western Europe is thinner, and small markets (Nordics, Gulf states, most of Africa, most of Latin America outside Brazil) may return zero usable free proxies for hours at a time.

If you need a specific country reliably, free lists will not get you there. That is exactly where our country pages help — we maintain per-country breakdowns of which providers actually hold IPs in each market, so you can see at a glance whether a geography is realistically servable for free or needs a paid pool.

The 7 real risks of free proxies (and how to neutralise each one)

Free proxies are not dangerous if you understand what you are doing. They are dangerous when people treat them like a VPN. Here is the honest risk list and the mitigation for each.

• Traffic interception — the operator sees every unencrypted byte. Mitigation: only ever send https:// through a free proxy, and never anything sensitive.
• TLS stripping / fake certificates — some free proxies downgrade or MITM HTTPS. Mitigation: never click through a certificate warning on a free proxy; treat it as proof the proxy is hostile.
• Content and script injection — ad or crypto-miner injection into plain HTTP pages. Mitigation: HTTPS-only, plus an ad blocker.
• Credential theft — the number-one way people get burned. Mitigation: never log into anything through a free proxy list entry. Not email, not social, not banking, not your scraping targets.
• Shared reputation — thousands of people used that IP before you, so it is likely already blocklisted. Mitigation: expect blocks; rotate aggressively; never judge a target's defences by free-proxy results.
• Vanishing uptime — free proxies die mid-session. Mitigation: short timeouts (5–8s), automatic rotation, idempotent requests.
• Legal and ToS exposure — many free proxies are misconfigured machines whose owners never consented. Mitigation: prefer explicitly-published free lists, keep volumes low, and read our disclaimers before scraping anything commercially.

Free proxy vs free VPN vs free trial: which free option fits your job

'Free' covers three very different products and picking the wrong one wastes days.

A free proxy list gives you volume and variety with zero signup, zero encryption guarantees and near-zero reliability — best for scraping public pages, rotating IPs cheaply and testing how a site behaves from other networks.

A free VPN encrypts your whole device but funnels you through a few heavily-flagged IPs, usually caps bandwidth, and in several documented cases monetised users' idle connections. Best for casual privacy on public Wi-Fi, useless for scraping.

A free trial or permanent free tier from a real provider gives you a small amount of clean, authenticated, geo-targetable bandwidth with actual uptime — the free option that behaves like a paid product. Webshare's permanent free tier (10 datacenter proxies plus free monthly residential bandwidth, no card required) is the single best free starting point if you need requests to actually succeed, and most major providers add free trial credit on top.

The real math: when free proxies start costing more than paid

This is the part no free proxy list site will tell you, so here it is with numbers. Measure cost per successful request, not cost per proxy.

A merged free proxy list typically yields a 5–20% success rate against a defended target, with median latency of several seconds and frequent mid-request failures. To land 100,000 successful requests at a 10% success rate you issue roughly one million requests, and each retry burns your compute, your bandwidth and your engineering attention. Add the checker infrastructure, the retry logic, the dead-proxy monitoring and the debugging time when your data silently goes half-empty, and a few developer hours a week is already more expensive than the entire proxy line item it was meant to avoid.

The same 100,000 requests through a residential pool at roughly $1.75–$2 per GB with a 95%+ success rate typically lands in the low tens of dollars — paid once, with no retry tax and no maintenance. The honest threshold: below about a thousand requests a day on public, undefended pages, free proxies are genuinely the right answer. Above that, or any time the data matters, free becomes the most expensive option you can choose.

The practical strategy most teams land on: build and prototype on free proxy lists, keep a free tier account for smoke tests, and route production through a paid pool. Compare current pricing across every provider we track on our comparison page before you commit to anything.

Free proxy list hygiene: 12 rules that make free proxies usable

Follow these and free proxies go from 'nothing works' to genuinely productive.

• Re-fetch your free proxy list at least hourly — anything older than an hour is mostly dead.
• Always merge multiple free sources; single-source lists overlap heavily and die together.
• Deduplicate by IP:port, then by IP — many free entries are the same host on several ports.
• Test before use, never during. Keep a verified pool, not a raw list.
• Use 5–8 second timeouts. A free proxy that takes 20 seconds is not slow, it is dead.
• Filter to elite when the target inspects headers.
• Match protocol to target: HTTPS-capable free proxies for https:// URLs, SOCKS5 for non-HTTP traffic.
• Rotate on every request, not every session.
• Cap concurrency per proxy at one or two connections — free proxies collapse under load.
• Retry across proxies, not on the same proxy.
• Log which free source produced the working proxies and drop sources that stop delivering.
• Never authenticate through a free proxy. Ever.

Keep this page bookmarked — free lists rot, this method does not

Free proxy websites go dark constantly; the ten sources above are the ones that have survived because they are automated and public. Bookmark this page, run the checker on a schedule, and you will always have a fresh free proxy list without trusting any single site.

When free stops being enough — and if your project is real, it will — start with a free tier so you pay nothing to find out whether a clean pool fixes your success rate, then scale into the cheapest provider that clears your target. Our reviews, per-country pages and scraper API comparison exist for exactly that transition.

Frequently Asked Questions