What no logs can honestly mean
No logs is the most repeated claim in the VPN business, and it is often stated more widely than it can be true. A VPN server cannot hold nothing. To send replies back to you, it has to know, at that moment, where you are.
So the claim worth making is narrower and more useful: nothing about what you do is written down, and nothing about your connection is kept after it ends. It helps to split logs into four kinds.
| Kind of log | What it would hold | On Veto |
|---|---|---|
| Activity | The sites you open, the name lookups, the content | Not collected |
| Connection history | When you connected, for how long, from which address | Not kept |
| Live session state | Where you are right now, so replies can reach you | In memory only, gone when you disconnect or the server restarts |
| Account | Who the account belongs to | A 15 digit number, a plan, a paid until date and a key per device. No name or email |
The three things our server must hold
While your device is connected, the WireGuard software on the server holds three things about it. We read them on the New York server on 15 November 2026.
| What is held | Why | Where | For how long |
|---|---|---|---|
| The address you connected from | To send your traffic back to you | Memory | Until you disconnect or the server restarts |
| The time of the last handshake | To know when to agree on fresh keys | Memory | The same |
| A count of bytes sent and received | Kept by the tunnel software as it runs | Memory | The same |
The server also holds one public key and one tunnel address for each device, in memory: the account system sends it the list, and a restart clears it until the list is sent again. Those are what join a device to the server. They say nothing about what the device does.
None of the three is copied, exported or written to a file. The WireGuard software keeps the current state of each device, not a history of it.
How we checked that nothing else is written
A no logs claim is only as good as the checks behind it. These are the checks we ran on the New York server on 15 November 2026.
| Part of the server | What could log | What we found |
|---|---|---|
| The firewall | A rule can write a line for every connection | The full rule set was read. No rule writes a log |
| The name resolver | It can keep every name it was asked | Query logging is switched off in its settings |
| The tunnel | It could keep past connections | It holds the current state only, as listed above |
| The website | A web server logs each visit by default | The web server's visit log is off. Each page view sends the team a short note, as the privacy policy lists. Failed requests write one line, deleted after 14 days |
We also recorded a page load from both sides of the tunnel to see what a watcher could read. That test is written up in our guide What does a VPN hide?.
What our checks do not prove
We would rather you read this here than find it later.
- No outside audit has been done. The checks above are ours. Nobody independent has repeated them.
- A server can be changed. What we checked on 15 November 2026 was true that day. Our rule is that the privacy policy changes before the servers do, and every change to it is listed there with its date.
- The app's code is not published. What the app does can be checked from the outside, as this site's tests do, but its code cannot be read.
- The checks above were run on the New York server. The Pro cities were tested for speed, place and name lookups, as the servers page shows.
Why the account matters as much as the logs
A log is only dangerous if it can be tied to a person. Most VPNs tie every account to an email address and a card, so any record they keep, by choice or by mistake, is a record about someone named.
Veto's account is a 15 digit number made at random by the app or the account panel. There is no email, name or password behind it. If a record ever existed, it would be about a number. Our guide Can a VPN be tracked? sets this next to the list of records US law lets police demand.
How to judge any VPN's no logs claim
- Look for a list, not a slogan. A serious policy names what is held, why, where and for how long.
- Check what it asks for at sign up. Each item it collects is an item it holds.
- Look for the hard part. A policy that admits what a server must hold while you are connected is more credible than one that claims nothing is held at all.
- Look for what is outside the company. An audit, open code or a public record of requests from authorities each add something a promise cannot.
- Check the dates. A policy should say when it last changed and what changed.
What a restart clears
Because the three live items sit in memory, a restart of the server clears them all. After a restart, the server asks the account system for its list of devices again, and knows the public key and tunnel address of each one and nothing else. Devices that reconnect start a fresh handshake, and the server learns their current address again.
The same is true when you disconnect. The tunnel software keeps a device's entry, but the address you came from is replaced the next time you connect from somewhere else, and nothing records where you were before.
What the account system holds
The account system holds one record per number. It is short on purpose.
| Field | Held | Not held |
|---|---|---|
| Identity | The 15 digit number | Name, email, phone, password |
| Plan | Free or Pro, and a paid until date | A history of past plans |
| Devices | One public key and one place on a server per device | Device names, models or a list of past logins |
| Payment | A random payment reference and the payment company's codes | The card and the email, which stay with the payment company |
Questions about no logs
What does a no logs VPN mean?
A VPN that does not record what you do through it: no list of sites, no name lookups, no history of when you connected. It cannot mean a VPN that holds nothing at all, because a server must know where you are while you are connected in order to send your traffic back.
Is any VPN truly no logs?
Every VPN holds some state while you are connected. The honest test is what is written down and kept afterwards. Look for a list of what is held, why and for how long, and for proof beyond a promise.
Does Veto keep connection logs?
No. Our New York server keeps no record of when you connected or for how long. While you are connected it holds the address you came from, the time of your last handshake and a byte count, in memory only.
Has Veto's no logs claim been audited?
No. No outside audit has been done. This page shows how we checked our New York server and says plainly what that does not prove.
Does the account record show when I used Veto?
No. The record has no date the account was made, no date it was last used and no list of connections. It holds the number, the plan, a paid until date and one key for each device.
Can a no logs VPN be forced to start logging?
A server can always be changed by whoever controls it. That is why a written policy, published changes and outside checks matter: they make a change visible. Veto's rule is that the privacy policy changes before the servers do.