Why public MTProto proxies die quickly
A public MTProto proxy from an open Telegram channel lives, on average, from a few hours to a few days. Less often, weeks. The reasons are always the same: either the IP overheats from mass traffic and gets into filters, or the owner shuts down the server because the load has become more expensive than they are willing to pay for.
If several proxies from your favorite channel suddenly stopped working today, this is normal behavior for public MTProxy servers, not “that’s it, the messenger has been shut down.” This is simply their life cycle.
Below: what exactly kills a public proxy, where to get new ones, and why your own MTProxy is always more stable than any public one.
In short
- Public MTProxy = one IP that many users connect to in a short time.
- An overheated IP gets into TSPU filters in days or hours.
- VPS providers block such IPs because of complaints and traffic overages.
- The owner can shut down the proxy when they get tired of paying for traffic.
- The same proxy can work on one operator and not work on another at the same time.
- Your own MTProxy on a personal VPS lives much longer: there is no one burning it with mass traffic.
What kills a public MTProxy
The life of a public proxy is short but predictable. Understanding what stage it is at means understanding whether it makes sense to wait or to look for a new one right away.
IP overheating
As soon as a proxy link appears in a channel with tens of thousands of subscribers, thousands of new connections hit the server within minutes. TSPU systems notice the anomaly, the IP comes under observation, and then enters the filter. After that, even if the load is removed, the IP remains “dirty” for weeks.
Registry and complaints
If the proxy is used at scale to bypass blocking, official demands will sooner or later arrive for it. The IP/subnet enters the registry of prohibited resources, and the operator is required to cut traffic to that address. With TSPU, this happens automatically.
The owner shut down the server
A VPS costs money. 1,000 active users can easily push terabytes per month. Many enthusiasts shut down a proxy when the bill starts to hurt or they simply lose the desire to maintain it. This is also a reason for “it suddenly stopped working.”
Complaint to the VPS provider
If complaints are filed against the IP (for example, for spam or phishing through the same subnet), the VPS provider may temporarily or permanently disable the server. This is especially common with large hosters that automatically react to abuse reports.
Secret changed
Sometimes a proxy does not “die,” but simply changes its secret after a wave of DPI filters. Old links stop working, and new ones are published in the same channel. If the channel is reputable, it is worth subscribing to notifications.
How to tell that a proxy is dead
| Symptom | What happened | What to do |
|---|---|---|
| The proxy does not respond on any device or with any operator | The server is shut down or the IP is blocked globally | Take a new one from a trusted channel |
| The proxy works abroad, but not in Russia | The IP is in the registry or is being cut by TSPU | Take a new one, preferably from a new subnet |
| The proxy works on one operator, but not on another | Partial IP blocking | Keep several proxies for different operators |
| The proxy works for 5–10 seconds, then drops | The server is overloaded | Wait for the load to fall or switch to another one |
| The proxy worked in the morning and died by evening | Daytime load finished off an overheated IP | Replace it with a new one |
| All proxies from the channel stop working at once | The channel is outdated or the owner lost access | Find another source |
How to check right now
The easiest way to tell “the proxy is dead” from “the operator is cutting off me specifically” is with a comparative check. If the same proxy works normally on another operator, the problem is filtering by your operator, and you need another proxy or another tool. If it works nowhere, the server is finished and there is no point waiting.
Freedom Checker is designed for exactly these checks using probes in different regions: you can see whether a specific service is available on specific operators right now. From indirect data, you can immediately understand whether your proxy has dropped out of the live ones.
Failure or blocking?
A public MTProxy failure looks like “it worked yesterday, not today.” Signs: one proxy is unavailable, the other proxies from the same channel are alive, other users in the chat complain “same for me.” Treatment: take another proxy.
Blocking looks different: different proxies die at the same time on a specific operator, or the proxy is accessible from outside Russia but not inside. Then this is not “the proxy broke,” but targeted filtering. You should change not one server, but the approach: your own MTProxy, or a VPN with obfuscation.
What to pay attention to
- Do not tie yourself to one proxy — keep 3–5 backups.
- Subscribe to channels that update lists regularly, not ones that post “once every six months.”
- Choose proxies with fake-TLS (the ee prefix in the secret) — they live longer than basic ones.
- The more popular the proxy source, the faster the IP will get filtered.
- Do not enter secrets manually — use tg://proxy links.
- If you regularly need Telegram, run your own MTProxy on a cheap VPS.
- Remember that a list of “10 working proxies” from a year-old article is 99% already dead.
Conclusion
A public MTProto proxy is a consumable, not permanent infrastructure. It is created to quickly give access to many people, and it dies just as quickly under its own load and filters. This is normal, and the only reliable way to bypass this cycle is your own server, known to no one except you and a few of your devices.
Before replacing the proxy, check whether this is a problem with that specific server or general filtering by your operator. The answer determines how many proxies you will have to try.