Does Discord IP ban? How bans work and how to stop ban evasion

When you ban a member from a server, two things happen:

1. The account is banned. That Discord account can no longer join the server, whatever invite it uses, until someone removes it from the ban list.
2. The IP address is banned. Other accounts connecting from the same IP address are prevented from joining too.

There is no setting for this. You do not need a command or a bot to “IP ban someone on Discord”: every ban from the member menu, from /ban or from a moderation bot is already account plus IP.

A few details are worth knowing:

• You never see the IP address. Discord does not show it to owners, moderators or bots, so you cannot look it up, compare it or ban it separately.
• A kick is not a ban. A kicked member can rejoin with any invite.
• A timeout silences a member for up to 28 days but leaves them in the server.
• A server ban is local. It has no effect on other servers. Only Discord's Trust & Safety team can suspend an account across the platform.

An IP address identifies a connection, not a person. Most people do not have a fixed one, and changing it takes no skill:

• Mobile data. Switching a phone from Wi-Fi to 4G or 5G gives a different address, often shared with thousands of other subscribers.
• A VPN or proxy. One click, a new address, usually in another country.
• A router restart. Many home connections receive a new address every time the modem reconnects.
• Another network. School, work, a friend's house, a café.

IP bans also hit the wrong people. A banned student shares a campus address with hundreds of others, and a mobile carrier can put a whole neighbourhood behind one address. So the IP ban is both too weak against the person you banned and occasionally too strong against people you did not.

Ban evasion is returning to a server you were banned from, using another account or another connection. The routine is almost always the same: create a new account (an alt account), change the IP address if needed, and rejoin with a public invite.

Discord treats this as a violation: its Community Guidelines list “ban or block evasion” under harassment. You can report an evader to Discord, and you should when the behaviour is serious. But a report is slow, it handles one account at a time, and it does nothing for the account created an hour later.

Signs that a new member is a returning one:

• The account was created the day of the ban, or shortly after.
• It heads straight for the same channels and the same people.
• The writing style, the grievances and the active hours match.
• It already knows your rules well enough to stay just inside them.

• Raising the verification level. Requiring a verified email or phone slows people down. It does not stop anyone with a second email address or a prepaid SIM, and it never compares the new account with the banned one.
• Captcha bots. A captcha proves that a human is present. The person you banned is human.
• Banning by username. Names cost nothing to change.
• Closing invites for good. It stops evaders and everyone else. It is a lockdown, not a moderation policy.

All of these act on things the evader controls. What they cannot easily change is the combination of device, browser and network they actually use.

A ban sticks when the server can recognise the person at the door instead of the account. In practice that means four layers:

1. Verify every new member before they can talk. One verification channel, one verified role, everything else locked. Our guide to setting up verification shows how.
2. Use alt detection at that gate. A verification bot such as Double Counter compares the device, browser and network signals of each new account with the accounts it has already seen, including the one you banned, and flags the match.
3. Block VPNs, proxies and Tor during verification. This removes the easiest way to look like someone new. Members can turn their VPN back on once they are verified. More on VPN blocking.
4. Hold very young accounts. A minimum account age of a few days makes “new account, same evening” impossible.

Then choose the consequence. For a server that deals with persistent evaders, auto-banning detected alts closes the loop without a moderator being awake. For most communities, a quarantine role plus a staff alert is a safer start, because a human looks at the evidence before anyone is removed.

Stronger enforcement needs a visible way back, otherwise people who would have appealed simply evade. Three habits keep things fair:

• Offer an appeal route outside the server, such as a form or a small appeals server, and say so in the ban reason.
• Log why each ban happened. When an alt is detected six months later, staff should be able to see what the original ban was for.
• Review detections before permanent action when you can. People who share a household or a device can look related, and staff can approve a legitimate second account by hand.