Best No-Logs VPN: What Happens After You Disconnect?
- “No Logs” Doesn't Always Mean the Same Thing
- Find Out What Survives Your Session
- Know Who You're Trusting With Your Traffic
- Check Whether the Audit Examined Logging
- RAM-Only Servers Answer One Part of the Question
- Compare Policies Without Reading Every Line Twice
- Test the App Separately From the Policy
- Choose Once the Privacy Answers Are Clear
A VPN can promise “no browsing logs” and still keep a record of when you connected and which IP address you used. Before you buy, find out what survives after you press Disconnect.
A convincing no-logs promise needs specifics: no saved browsing activity, no connection records that identify you, and independent checks that actually examine those claims. You shouldn't need a free weekend and a magnifying glass to find the answers.
You can cut through that fine print by asking what each record reveals about you. A VPN can remember that you paid the bill without remembering where you went online.
“No Logs” Doesn't Always Mean the Same Thing
A log is a record of an event, such as a connection starting or an app crashing. The privacy question is what that record contains and whether anyone can connect it to you.
A subscription receipt shows that you paid for a VPN. A record linking your home IP address, connection time, and browsing destination could reveal part of your online activity.
Those details often sit in different sections of a policy, so look beyond the paragraph labeled “no logs.” Use these four categories to keep your comparison manageable.
| Type of data | What it can include | What to look for |
|---|---|---|
| Browsing activity | Sites visited, DNS requests, unencrypted traffic contents | Explicit statements that these aren't logged |
| Connection details | Your IP address, connection times, server used | Which details are processed, which are saved, and deletion rules |
| Account records | Email address, payments, support messages | What identifies you and how long it's kept |
| App diagnostics | Crash reports, device identifiers, usage analytics | What reports contain, who receives them, and whether they're optional |
DNS requests are the lookups that translate website names into addresses computers can use. They deserve attention because a saved lookup can reveal a domain you requested, even without a record of the page's contents.
Keeping account information doesn't automatically contradict a no-activity-logs promise. For example, Mullvad's policy separates its exclusions for traffic and connection logs from the information it handles for payments and support.
The useful distinction is whether retained records could link your identity to your VPN activity, including when records from separate systems are combined.
Find Out What Survives Your Session
A working VPN needs temporary information to carry traffic between your device and its servers. WireGuard's protocol documentation, for instance, describes connection state and encryption keys that change during use.
That doesn't tell you whether a VPN company saves connection records elsewhere. The protocol and the provider's logging practices answer different questions.
Consider two hypothetical services that don't save the websites you visit. One discards identifying connection details when they're no longer needed for your session; the other stores your home IP address and connection times for a month.
The second service retains a history of when you used it, even though it has no browsing history. Ask what remains after disconnection, how long it remains, and what it can be linked to.
Look for a specific deletion period or event. If a policy says “as long as necessary,” check what the information is needed for and what triggers its removal.
“Anonymous” needs a closer look, too. Removing your name isn't enough if a stable device identifier still lets the company join records from one visit to the next.
Know Who You're Trusting With Your Traffic
Those records matter because using a virtual private network (VPN) changes who handles your connection. Traffic covered by the VPN travels through an encrypted tunnel to its server, and websites see the server's public IP address instead of yours.
Your internet provider can still see that you're connecting to that server. The VPN operator handles the onward traffic, although HTTPS, the encryption used by secure websites, continues to protect page contents in transit.
The Federal Trade Commission's VPN guidance recommends researching the developer and reviewing the app's permissions before downloading. A request to read your text messages, for example, deserves an explanation.
Find the company name and any parent company, then check which outside services receive customer data. You need to know whose policy you're relying on before you can judge the promise.
Check Whether the Audit Examined Logging
Once you've read the promise, look for evidence that someone checked how it works. An “independently audited” badge is a starting point, but the report's scope tells you what the auditors actually examined.
A security test of a phone app might find software flaws without investigating server logs. An assessment of VPN servers might leave out account systems or analytics elsewhere in the service.
For a concrete example, Cure53's 2023 Mozilla VPN report describes testing client apps across five operating systems. It illustrates why an app assessment shouldn't be treated as proof of a service-wide no-logs claim.
When reading an audit, check:
- Scope: Did the work examine logging practices and relevant infrastructure, or a narrower part of the product?
- Date: When did the inspection happen, and have major systems changed since then?
- Findings: What problems or limitations did the auditors identify?
- Follow-up: Were fixes independently checked, or only announced by the provider?
A detailed report can be more useful than a spotless-looking badge. You can see what was examined, what needed work, and whether the provider followed through.
Still, an audit describes a particular assessment, not every future version of the service. The Electronic Frontier Foundation's VPN guide cautions that practices can change afterward.
Treat accounts of court proceedings or server seizures with the same attention to scope: check which records were sought and what was available in that case. Open-source apps provide another useful window into the software, but their public code doesn't expose everything running on private servers.
RAM-Only Servers Answer One Part of the Question
Audits may mention RAM-only servers, which run using memory that loses its contents when power is removed. ExpressVPN describes this approach as preventing its server software from writing data to a hard drive and clearing memory on reboot.
That reduces what can remain on the machine, but it doesn't establish what the whole company keeps. A server could send information to another system while it's running.
Look for explanations of the systems handling logins, DNS, performance monitoring, and support. Where data goes matters as much as where it starts.
Connection limits offer a useful example of the detail a provider can explain. Mullvad says its servers check a central service to validate accounts and simultaneous connections, with that information handled in temporary memory rather than stored permanently on disk.
That's a specific operational claim you can compare with audit coverage. A bare “RAM-only” label gives you much less to work with.
Compare Policies Without Reading Every Line Twice
With those distinctions in mind, open the policies for two or three providers you're considering. Search each page for “IP address,” “retain,” “delete,” “analytics,” and “third parties,” then read the surrounding paragraphs.
For each provider, write down three answers:
- What survives a session? Include identifying connection details, account records, and diagnostics, with their retention periods.
- What supports the promise? Note a relevant audit's date and scope, plus any unresolved findings.
- Who controls the data? Record the operator, parent company, and outside recipients named in the policy.
Mark an unanswered point as “unclear” rather than treating silence as a promise. That makes gaps visible when you compare the providers side by side.
If one of those gaps could change your decision, ask support something specific: “Do you store my source IP address after I disconnect, and if so, for how long?” A useful answer should identify the practice and point to the policy that covers it.
Include the company's location in your comparison, but don't use a country label as a substitute for checking its records. Revisit the policy after an ownership change or a major product update, since a familiar app name can outlast the arrangements behind it.
Test the App Separately From the Policy
A convincing policy still needs an app that works on your devices. Leak tests can reveal unexpected exposure of your public IP address or DNS requests, but they can't show what a provider stores behind the scenes.
Check both IPv4 and IPv6, the two versions of internet addressing, if your connection supports them. Review the kill switch as well: it's intended to block traffic when the VPN connection fails.
Use the provider's testing instructions, since deliberately clicking Disconnect may behave differently from an unexpected interruption. The result should match the protection the app promises in that situation.
Review split tunneling, which lets selected traffic bypass the VPN, and any custom DNS settings. A route you've deliberately excluded needs to be considered separately when you interpret a leak test.
Even with the app working correctly, signing into a website identifies you to that site. Cookies and browser fingerprinting, which recognizes patterns in your browser's characteristics, can also allow tracking despite a changed IP address, as the EFF explains.
That wider picture of data collection is the subject of Privacy Is Power. It provides context for the privacy choices that remain after you've chosen a VPN.
- Connects everyday data collection to real choices about freedom, power, and control
- Explains why privacy matters even when you have nothing to hide
- Turns a broad social issue into practical questions you can apply to your digital life
Choose Once the Privacy Answers Are Clear
Use the policy and supporting evidence to narrow your shortlist, then compare device support, everyday performance, and subscription terms. Privacy claims won't help much if the app is so awkward that you keep turning it off.
Before paying, you should be able to explain what the provider retains and why you trust its account of that. If a key answer is still missing, resolve it before committing to a long subscription.