Free proxies for YouTube fail at the thing people want most, which is video. Video needs speed that holds. YouTube Help recommends a sustained 2.5 Mbps for 720p and 20 Mbps for 4K. On 28 September 2026 we timed 200 working free proxies on ordinary addresses from our own list, twice. 35 and 30 kept up the speed for 720p, and 1 of the 400 reached the speed for 4K.
This page covers the test, what the YouTube terms say, and why your Google account should never pass through a free proxy. It then shows the little a free proxy is good for on YouTube.
Do free proxies for YouTube actually work?
For a page, sometimes. For video, rarely. A video plays well only when the connection keeps up with it, second after second. YouTube Help gives the numbers, as the "Recommended sustained speed" for each quality.
| Video quality | Recommended sustained speed |
|---|---|
| 4K UHD | 20 Mbps |
| HD 1080p | 5 Mbps |
| HD 720p | 2.5 Mbps |
| SD 480p | 1.1 Mbps |
| SD 360p | 0.7 Mbps |
A free proxy has to pass every byte of the video along. So the question is simple: how fast is a free proxy, and does it hold that speed?
We timed 200 free proxies on a 3.3 MB download, twice
We picked 200 entries at random from the working part of our free proxy list, only from addresses that held fewer than 20 of them, and 48 minutes later another 200. All had passed our HTTPS check, since YouTube runs on HTTPS. Through each one we downloaded the same 3.3 MB file from our own website, with 30 seconds in total. We did not test against YouTube itself, because its terms forbid accessing the service "using any automated means". We then compared each speed with the table above.
- 35 of the 200 held at least 2.5 Mbps, the speed for 720p. 16 of them reached the 5 Mbps for 1080p, and none the 20 Mbps for 4K. In the second run, 30 held 720p, 14 held 1080p and 1 held 4K.
- 65 delivered the file, whole or still arriving when the 30 seconds ran out. Their median speed was 2.95 Mbps, and the fastest reached 13.2 Mbps. In the second run, 49 did, at a median of 3.14 Mbps.
- 104 sent nothing within 30 seconds, 20 had no working connection, and 11 presented a certificate our client refused.
- Counting every entry of the list, our two samples of 100 on 27 September found 22 of 200 that held 720p, and a median of 1.9 Mbps among those that delivered.
Straight from our server, the same file came at 6.6 to 30.9 Mbps during the two runs, which shared the server with our other tests that evening. On 27 September, with the server to itself, it came at 21 to 310 Mbps.
So about one free proxy in six could play YouTube at 720p, and 1 of 400 at 4K. Most could not play a video at all.
What the YouTube rules say
The YouTube Terms of Service, dated 15 December 2023, do not mention proxies or VPNs. Three of its rules still matter here. You may not do the following.
- "circumvent, disable, fraudulently engage with, or otherwise interfere with any part of the Service (or attempt to do any of these things), including security-related features or features that (a) prevent or restrict the copying or other use of Content or (b) limit the use of the Service or Content".
- "access the Service using any automated means (such as robots, botnets or scrapers)", except for public search engines under its robots.txt file or with its written permission.
- Dodge a strike with another channel: "If your channel has been restricted due to a strike, you must not use another channel to circumvent these restrictions".
None of these depends on the address you connect from, and no proxy makes them go away.
Your Google account is worth protecting
Your YouTube login is your Google account, with your email and maybe a channel behind it. A proxy is a server run by a stranger, in the middle of your traffic. It adds no encryption of its own. YouTube and Google sign-in run on HTTPS, so on a clean connection your password is encrypted in transit. A hostile proxy can still present a fake certificate and hope you click past the warning. A 2018 study of more than 107,000 listed open proxies found TLS man-in-the-middle attacks. In our two runs, 11 and 8 of the 200 free proxies presented a certificate our client refused.
Never sign in to YouTube, or to anything under your Google account, through a free proxy. Some proxies also pass your address on. On 28 September 2026, 23.9 percent of the working web proxies on ordinary addresses in our pool were transparent, and 5.1 percent counting every entry. The site at the other end saw the address of the person using them. Our guide to whether free proxies are safe covers the rest.
Where free proxies for YouTube are fine
Free proxies suit anonymous, throwaway, browser-only tasks. A dead proxy, or a watched one, costs you nothing there. On YouTube, that leaves little.
- Opening a public YouTube page in a browser, signed out, to see that a proxy works at all.
- Learning how a proxy works in a browser, with a throwaway request.
None of these involve your Google account or a video you need to watch. The moment one of those enters the picture, a free proxy is the wrong tool.
| What you want to do | Free proxy? | Why |
|---|---|---|
| Watch in HD or 4K | No | 16 and 14 of 200 held 1080p, and 1 of 400 held 4K. |
| Watch at all without buffering | Rarely | 35 and 30 of 200 held the speed for 720p. |
| Sign in to YouTube or Google | No | A hostile proxy can try to read your login. |
| Download or scrape videos | No | The terms forbid automated access and downloads without permission. |
| Dodge a strike or a ban | No | The terms forbid using another channel to get around a strike. |
| Open a public page, signed out | Sometimes | A web page with no login, while the proxy stays up. |
Where a free proxy costs you
Anything you want to last is the wrong job for a free proxy. Most of our 200 sent nothing at all in both runs, and the ones that did ran at a median of 2.95 and 3.14 Mbps. For web work that has to keep running, a residential proxy is the better tool. Its address belongs to a home connection, and it stays up while you work. Ours start at $0.44/GB, pay as you go, and purchased traffic has no scheduled expiry date. Video is heavy, though: at the 5 Mbps YouTube recommends for 1080p, one hour comes to about 2.25 GB through any proxy.
Test any proxy before you use it
A free proxy that worked ten minutes ago may already be gone. Two things decide whether it is usable: whether it is alive, and where it actually exits. Paste any candidate into our free proxy checker, with no signup. It makes a real connection through the proxy and reports the exit address, country, latency and anonymity grade. From the command line, the same check is one request to our free checker API.
# Tests one proxy from our server: alive or dead, protocols, anonymity, latency, country
curl "https://hproxy.com/api/proxy-check?proxy=203.0.113.7:1080"
A proxy that answers the check can still be too slow for video, as our test showed. The full method is in how to check if a proxy is working. It also shows how to spot a proxy that leaks your address in the headers.
The plain answer
Free proxies for YouTube cannot carry what YouTube is for. On 28 September 2026, 35 and 30 of 200 working free proxies on ordinary addresses held the 2.5 Mbps YouTube recommends for 720p. 16 and 14 held 1080p, and 1 of 400 held 4K. Most sent nothing at all within 30 seconds. The YouTube terms forbid getting around features that limit the use of the service, and your Google login should never pass through a stranger. Keep free proxies to throwaway checks, signed out. For web tasks that can fail, start with our free proxy list, and check anything you find elsewhere with the proxy checker.
How we measured
Both runs went from our server in the United States on 28 September 2026 and finished at 22:29 and 23:17 UTC. Each took 200 working entries of our free pool at random from addresses that held fewer than 20 of them, all past our HTTPS (CONNECT) check. Each result records how many working entries shared its address, never the address. Each proxy first fetched a test address of our own, then downloaded the same 3.3 MB file from our website. It had 30 seconds in total, with compression turned off so that it could not inflate the speed. The speed is the part of the file that arrived, divided by the time from its first byte to its last. We never contacted YouTube. The figure over every entry comes from two samples of 100 of the whole pool on 27 September, between 20:38 and 20:55 UTC. The transparent share comes from the working web proxies in our pool at 18:28 UTC on 28 September. We store no proxy addresses from these tests, and every time is in UTC.
Sources
- YouTube Help, System requirements & supported devices for YouTube, the table of recommended sustained speeds, read on 27 September 2026.
- YouTube, Terms of Service, dated 15 December 2023, read on 27 September 2026.
- Akshaya Mani, Tavish Vaidya, David Dworken and Micah Sherr, An Extensive Evaluation of the Internet's Open Proxies, arXiv 1806.10258, 27 June 2018.
- HProxy documentation: the proxy checker API, read 27 September 2026.
