How to Prevent Discord Raids

A raid is a coordinated wave of accounts joining a server in a short window to spam, post offensive content, or overwhelm moderators before anyone can react. Raids are usually automated or semi-automated, which is why they can add dozens of accounts in seconds — far faster than a human moderator can manually review and remove them one by one.

Public servers, servers with an easily-found invite link, and servers going through a sudden spike in attention (a viral post, a leaked invite) are the most common targets.

Discord offers a server verification level and a raid-protection toggle that can temporarily restrict new members. These are blunt, server-wide controls — genuinely useful in the middle of an active raid, but they don't distinguish a suspicious account from a legitimate new member, and someone on staff has to notice the raid and turn them on manually.

Anti-Raid watches your server's join rate continuously, so it doesn't need a staff member to notice anything first.

• Free on every server: fixed-rate detection — 5 joins in 10 seconds — flags a likely raid and warns staff, without taking automatic action.
• Server Premium makes the threshold and detection window fully configurable, and adds more signals: default-avatar accounts, likely bot accounts, clusters of similar usernames, and account age.
• Premium can also act automatically — kick, ban, or quarantine the accounts it flags — and posts a guarded alert so staff can resolve the incident, prune the detected members, or keep protection active from one place.
• Rollback support restores the protection state Anti-Raid was in before the incident, so a defensive setting doesn't stay clamped down longer than it needs to.

1. Add Airwavy to your server — free-tier join-rate detection is active immediately, with no configuration required.
2. If you have Server Premium, open Anti-Raid in the dashboard and review the join threshold and detection window against your server's normal traffic — a large server with frequent invite campaigns needs looser defaults than a small private one.
3. Turn on the detection signals that fit your community; default avatars and similar usernames are a good starting point for most servers.
4. Choose a private alert channel and the staff roles allowed to act on alerts.
5. Decide the automatic response — start with alert-only or quarantine while you confirm normal traffic doesn't trip it, then move to automatic kick or ban once you trust the configuration.

• Setting thresholds so tight that a normal event — a giveaway, a partnered server's shoutout — gets flagged as a raid. Validate against your server's actual traffic before tightening.
• Jumping straight to automatic ban. Start with alert-only or quarantine, confirm detection is accurate for your server, and only then enable stronger automatic actions.
• Confusing Anti-Raid with Anti-Nuke. Anti-Raid watches new joins; Anti-Nuke protects against a compromised staff or integration account making rapid destructive changes, like mass channel deletion or mass bans — they solve different problems, and most servers want both.
• Forgetting that a compromised owner account is the one gap automation can't close. Anti-Raid and Anti-Nuke protect the server; Discord's own account security — 2FA, trusted devices — protects the owner account itself.