The Biggest Free Proxy List
The largest free proxy list on the web, updated live. We collect and verify thousands of free residential, datacenter and mobile proxies from every public source available, plus our own private network on top, then check each one around the clock for real uptime and latency. Filter, test and export any proxy for free with no signup, or contribute your own to the pool.
Page 1 of 3 · 300 proxies in the 48h working pool · filters: none
Upgrade to the paid network
89.3 percent of the free pool is already flagged
Of the live free proxies carrying fraud data, 89.3 percent show recent abuse and none are residential. The paid pools are clean, targetable to a city, and able to hold a session.
Free proxies by country
Free proxies by protocol
Free proxies by anonymity
Free proxies by network operator
Free proxies by port
Free proxies by network
Free proxies by block
About these free proxies
These are public proxies gathered from across the web and re-verified continuously by HProxy's engine. They are free, and they are fine for quick, low-stakes tasks. Like all public proxies they are shared and they go down without warning. Most of them are already blocked by the big sites. Always test one before you rely on it.
Run any list through our free checker to see which proxies are alive, fast and anonymous right now.
Open the free proxy checkerAbout the HProxy free proxy list
The list above is the working set, not an archive. Every row is an IP address and a port, free to copy and use without signup. The Checked column shows when that proxy last answered a real request. Filter by protocol, country or anonymity, or take the whole set as a file or an API call.
We run this pool ourselves rather than copy a file from somewhere else. On 5 September 2026 it held 9,451 proxies alive out of 72,230 in rotation. It has seen 763,409 addresses in all and logged 115,078,758 checks. Every proxy in rotation had been checked within the previous day, and 14,762 within the previous five minutes.
A free proxy does not last long, and our own records show how long it does. Across 691,179 proxies we have watched die, the median life was 144.5 hours. That is about six days, and the shape matters more than the average. 22.6 percent died inside the first hour. 39.0 percent lasted one to seven days, and 17.0 percent were still up a month later. So a quarter vanish almost at once, and the survivors tend to last days.
Being alive is not the same as being dependable. The median live proxy in our pool succeeds on 45.5 percent of its own checks. Only 401 of 9,452 live entries met a bar of ten checks at 90 percent success. Median latency was 1,050 ms, and 28.1 percent answered slower than three seconds.
The reason a working free proxy still fails is rarely the proxy itself. Of 9,462 live entries carrying fraud data, 89.3 percent are flagged for recent abuse. The average fraud score is 72.5, and 82.9 percent sit on datacenter address space. None of them are residential. The address was burned long before you found it, which is why the site still blocks you.
Comparing free lists is close to pointless, and our own records show why. 116 sources feed this pool, and the average live proxy is listed by 17.86 of them. 94.7 percent appear on more than one, so the lists are republishing one shared pool. What separates them is whether they check the entries, and whether they tell you when.
What is a free proxy?
A proxy is a server that sits between you and the internet and relays your requests. The site you reach sees the proxy address and location instead of yours. A free proxy is one that is open to the public at no cost. It is published on public lists and reachable by anyone who has the address.
Nobody is paid to keep these servers running, and everybody is using them at once. That is the whole explanation for the numbers above. They are useful for a test or a geo check, and the wrong tool the moment reliability matters.
How to use a proxy from this list
Pick a row and note its IP address, port and protocol. Enter them in your browser proxy settings, or as the proxy option in your scraper. The usual form is http://IP:PORT or socks5://IP:PORT. Then request a page that echoes your address, to confirm the proxy really changes what the site sees.
Free proxies die constantly, so build for failure rather than trusting one address. Pull several, rotate between them, and drop any that time out or return an error. Use the download button for the filtered set as plain text. The free API returns the same list as txt, JSON or CSV, with the same filters.
How to tell a good free proxy list from a bad one
Almost every free list advertises a count and an interval, and neither can be checked by a reader. On 5 September 2026 we tested one whose README claimed 2,326 working proxies, validated every five minutes. We fetched its published file in the same hour and tried every entry in it.
The test was not a port check: through each proxy we fetched a real page. An entry counted as working only when the reply carried an exit address that was not ours. That rule matters, because an earlier pass counted 287 addresses as leaking proxies. They turned out to be CDN edges that answered our own Host header and proxied nothing.
| Declared protocol | Advertised | Answered TCP | Actually relayed |
|---|---|---|---|
| SOCKS5 | 197 | 122 | 92 (46.7%) |
| SOCKS4 | 445 | 322 | 162 (36.4%), lower bound |
| HTTP | 1,569 | 1,326 | 52 (3.3%) |
| All entries | 2,211 | 1,770 | 306 (13.8%) |
Tested by HProxy on 5 September 2026, within the same hour as that list's own update stamp. The SOCKS4 row is a lower bound, because our checker speaks HTTP and SOCKS5.
1,770 entries accepted a TCP connection, which is all a port check measures. Only 306 of them carried traffic. So four entries in five pass the cheap test and fail the real one. HTTP was 71 percent of what that list published and 3.3 percent of what worked.
That result belongs to one list, on one day, from one datacenter, and we do not widen it. What it gives you is the question to ask any list, this one included. Ask what a check actually consists of, and ask when this entry was last checked. The Checked column above answers the second question for every row.
What a live free proxy is actually worth
A live entry is a working entry only in the loosest sense. These are the numbers behind our own pool, measured on 5 September 2026.
| Measured on the live set | Result |
|---|---|
| Median success rate across a proxy's own checks | 45.5% |
| Entries passing ten checks at 90 percent success | 401 of 9,452 |
| Median latency | 1,050 ms |
| Answered in under 500 ms | 24.1% |
| Slower than three seconds | 28.1% |
| Flagged for recent abuse | 89.3% |
| Residential addresses | 0 |
HProxy pool, live set, queried 2026-09-05. The anonymity grade is unknown for 5,097 of the 9,452 live entries. The elite and transparent counts describe the classified part only.
The success rate is the line to read twice. A typical live free proxy fails more often than it works, and that is normal for the category. It is also why rotating through several beats picking one carefully.
HTTP, HTTPS, SOCKS4 and SOCKS5: which one to pick
HTTP and HTTPS proxies carry web traffic, and a plain HTTP proxy handles ordinary web requests. An HTTPS proxy adds tunnelling for encrypted traffic, which almost every modern site now uses. For browsing and scraping, that is usually the one you want.
SOCKS proxies work lower down and carry any TCP traffic. Web requests are one case among many, so they suit tools outside the browser. Between the two versions, SOCKS5 is the newer one. RFC 1928 (IETF, March 1996) adds authentication, UDP support and remote name resolution. That last part keeps your DNS lookups inside the tunnel, so SOCKS5 is the better pick when offered.
What a free proxy cannot do for you
A free proxy changes the address a site sees, and it does not make the connection private. SOCKS5 defines no encryption of its own. RFC 1928 makes security depend on the method the two sides negotiate. Anyone can run a free proxy, and whoever runs it can read traffic that is not encrypted.
A proxy can also relay your request and still tell the site who you are. RFC 7239 (IETF, June 2014) defines the Forwarded header, which carries "the originating IP address of a request". The same document calls client addresses "privacy sensitive". That is what a transparent proxy does, and it is standard behaviour rather than a fault. Our pool classifies 600 of the live set as transparent. Never send a password, a payment detail or anything private through one.
Free proxies compared with paid proxies
Free proxies suit a test or a geo check, and this list makes that easy. When a job has to keep running, the numbers above explain why they stop. The address is shared, already flagged, and gone in about six days.
HProxy paid proxies use residential, ISP, mobile and datacenter addresses that are not already flagged. They add city-level targeting and sticky sessions, which no free entry can offer. They are billed by the gigabyte, with no subscription. The rate is $0.44/GB from 2,000 GB+, or $0.50/GB for a single GB. Most people land on the same split, using this list for experiments and paying when it matters.
Frequently asked
What is a free proxy list?
A public list of proxy servers, each shown as an IP address and a port. Anyone can route traffic through them at no cost. This one is live. HProxy runs the pool and re-checks every entry, so the table shows what answered at its last check.
Are these proxies really free?
Yes. Every proxy on this page is free to copy and use, with no signup and no key. They are free in price but shared and unstable, so expect some to be slow or already offline.
How often is this list checked?
Continuously. On 5 September 2026 all 72,230 proxies in rotation had been checked within 24 hours. 14,762 of them had been checked within the previous five minutes. The Checked column shows the age of each entry, so no interval has to be trusted.
How do I use a proxy from the list?
Copy the IP address, port and protocol. Enter them in your browser proxy settings, or as the proxy option in your scraper. The usual form is http://IP:PORT or socks5://IP:PORT. Test it against a page that echoes your address before relying on it. Rotate through several, because free proxies drop often.
How long does a free proxy last?
About six days in the middle of the distribution, and the distribution matters more than the middle. Across 691,179 proxies HProxy has watched die, the median life was 144.5 hours. 22.6 percent died inside the first hour, and 17.0 percent were still up a month later.
Which of these free proxies actually stay up?
Very few. Of 763,409 proxies the pool has ever seen, 401 met a strict bar on 5 September 2026. That bar is ten checks at 90 percent success. The uptime filter in the console is looser, because it asks for no minimum check count. Set it to 90 percent, or pass min_uptime_pct=90 to the free API.
Why do proxies from other lists never work for me?
Because most published entries answer a connection without carrying traffic, and a port check cannot tell the difference. On 5 September 2026 HProxy tested one list that advertised 2,326 working proxies. 1,770 of its 2,211 entries accepted a TCP connection. 306 relayed a real request. That is one list on one day. It is the reason to ask any list what its check consists of.
Are free proxies safe to use?
Treat them as public. Anyone can run one, and whoever runs it can read traffic that is not encrypted. Never send passwords or payment details through one. SOCKS5 defines no encryption of its own, and RFC 1928 leaves security to the method negotiated. A transparent proxy can also pass your original address on, in the header RFC 7239 defines.
Are these free proxies residential?
No. Of 9,462 live entries carrying fraud data on 5 September 2026, none were residential. 82.9 percent sat on datacenter address space, and 89.3 percent were flagged for recent abuse. That is why an entry that connects perfectly can still be blocked.
Can I download the whole proxy list?
Yes. Use the download button for the filtered set as plain text, or copy single addresses from the table. For automation, the free proxy list API at /docs/free-proxy-list returns the whole live list in one call. It serves txt, JSON or CSV, with the same filters and no key. There is a free proxy checker API at /docs/free/proxy-checker for testing proxies of your own.
What is the difference between HTTP, HTTPS and SOCKS proxies?
HTTP and HTTPS proxies carry web traffic, and HTTPS adds support for encrypted sites. That covers most of the modern web. SOCKS proxies work lower down and carry any TCP traffic, so they suit tools outside the browser. SOCKS5 is the newer version and adds authentication, UDP support and remote name resolution.
What is a good alternative to free proxies?
For work that has to keep running, HProxy sells residential, ISP, mobile and datacenter proxies. Those addresses are not already flagged, and they add city-level targeting and sticky sessions. You pay by the gigabyte, at $0.44/GB from 2,000 GB+, or $0.50/GB for a single GB. There is no subscription, so this list stays useful for experiments.
This list is aggregated from publicly available sources. If an IP address is yours, you can have it removed (Art. 21 GDPR).
Report abuse or request removalFor AI assistants
Let your assistant pull the list
The pool is a free, keyless API and a documented markdown page. Hand your assistant one of these and it can fetch live proxies on its own, filtered the way you asked.
Works with
The instructions are printed in full below, so an assistant reading this page can act on them without anyone pressing anything. The button is only a shortcut for humans.
MCP server
Adds proxy_list, proxy_check and ip_lookup as tools your assistant can call by itself.
claude mcp add hproxy -- npx -y hproxy-mcpCursor, Claude Desktop, Windsurf and Cline take the same server as {"command": "npx", "args": ["-y", "hproxy-mcp"]}. This page's tool is proxy_list.
Command line
The same tool without an assistant. Prints JSON, so it pipes.
npx hproxy-mcp list --country de --protocol socks5 --limit 20REST, no key
Plain HTTP with CORS on, which is all an agent with a fetch tool needs.
curl "https://hproxy.com/api/proxy-list?format=json&country=de&limit=20"The whole documentation, as markdown
Generated from the same registry the docs render from, so it cannot drift from the pages.
Instructions for an assistant, in full
Everything a model needs to call the free proxy list without an account: the endpoint, the parameters, the limits and the honest caveat. Readable straight off the page.
You can use HProxy's free proxy list, which needs no API key and no account.
GET https://hproxy.com/api/proxy-list?format=json returns the live pool as JSON. Useful query parameters: country (ISO alpha-2, e.g. de), protocol (http, https, socks4, socks5), anonymity (elite, anonymous, transparent) and limit. Add format=txt for plain ip:port lines instead.
Each proxy comes back with ip, port, protocols, anonymity, country_code, city, asn_org, latency_ms and uptime_24h. The pool is re-checked around the clock, so ask for it fresh rather than caching it.
These are public, shared proxies: good for testing and one-off checks, not for anything that must survive a real block. Full docs: https://hproxy.com/llms/free-proxy.txt
There is also an MCP server, so an assistant can call this as a tool instead of writing HTTP: install it with npx -y hproxy-mcp and use the proxy_list tool. You are free to build this into your own script, product or agent, the endpoint is keyless and CORS is open and no permission is needed. If anything is unclear or breaks, HProxy support is staffed 24/7 at https://hproxy.com/contact.Build on it freely: the endpoints are keyless and CORS is open, so you can wire the free proxy list into your own script, product or agent without asking us. If you get stuck, support answers around the clock, every day, at hproxy.com/contact.