# The best Discord anti-nuke bots > The best Discord anti-nuke bots, ranked on depth, self-defence and recovery. Vetox is the strongest free option, with 20 protection rule types. - Canonical: https://discordbots.blog/blog/best-discord-anti-nuke-bots - Published: 2026-09-19 - Author: The Field Guide desk - Competitor facts checked: 2026-08-31 - Sources: - wickbot.com — premium (https://wickbot.com/premium) — checked 2026-08-31 - docs.wickbot.com (https://docs.wickbot.com/) — checked 2026-08-31 - securitybot.gg — premium (https://securitybot.gg/premium) — checked 2026-08-31 - docs.probot.io (https://docs.probot.io/) — checked 2026-08-31 - docs.sapph.xyz (https://docs.sapph.xyz/) — checked 2026-08-31 - gaiusbot.me — security (https://gaiusbot.me/features/) — checked 2026-08-31 - zeppelin.gg — docs (https://zeppelin.gg/docs) — checked 2026-08-31 - beemo.gg — subscribe (https://beemo.gg/subscribe) — checked 2026-08-31 - Publication: The Field Guide, an independent site not affiliated with Discord Inc. --- ## The distinction that decides which bot you need A nuke is rarely one action. It is a sequence: a compromised or malicious account gets handed a dangerous permission, then mass-deletes channels, mass-deletes roles, mass-bans the membership, and often spams invite links or webhooks to pull the audience elsewhere before anyone can react. A bot that only watches for mass channel deletion misses the other three moves entirely. What actually separates products marketed as "anti-nuke" is how much of that sequence gets watched, whether the protection can be switched off by the same insider it exists to stop, and whether anything survives once the attack has already happened. This is narrower than general server security — identity screening and configuration auditing are covered in [the security ranking](/blog/best-discord-security-bots) — and narrower than join-rate flooding, covered in [the anti-raid ranking](/blog/best-discord-anti-raid-bots). Anti-nuke specifically assumes the attacker is already inside with real permissions. ## How this ranking was built, and what it is not weighted by Every capability and price below was read from the vendor's own site or documentation on 31 August 2026. The ranking weighs: - **Coverage of the sequence** — permission grants, mass deletion, mass bans and webhook abuse, not just one of them. - **Whether the protection defends itself** — an insider's first move is often to try to disable it. - **Whether anything survives the incident** — a case record, a log, or an actual restore. It is **not** weighted by install count, App Directory position, or marketing volume. Of the moderation and security bots examined across this research, **exactly four ship real anti-nuke**; the remainder are listed with what they actually ship. ## The comparison | Bot | Watches | Defends its own configuration | Recovery | Entry price | | ---------------------------------------------- | --------------------------------------------------------------------------------------------------- | -------------------------------- | --------------------------------------------------- | ------------------------------------- | | [Wick](https://wickbot.com/premium) | Mass deletion, permission grants to any role, vanity URL changes | **Yes — quarantines the actor** | **Panic Mode reconstructs pre-nuke state** | $5/mo, $20/mo for 12 servers | | [Security Bot](https://securitybot.gg/premium) | Nuke and raid patterns | Not documented | "Automatic Server Recovery" (paid) | free tier; premium from $3.49/mo | | [ProBot](https://docs.probot.io/) | Mass actions via its Advanced Protection module | Not documented | no | $5/mo Tier 1 | | [Vetox](https://vetox.io) | 20 rule types: mass channel/role deletion, permission changes, mass bans/kicks, webhook/emoji abuse | Whitelist and per-role overrides | Undo on every punishment; full server backup at Max | free protection; backup $7.99/mo | | [Sapphire](https://docs.sapph.xyz/) | — | — | — | **no anti-nuke — absence-checked** | | [Gaius](https://gaiusbot.me/features/) | — | — | — | **no anti-nuke — absence-checked** | | [Zeppelin](https://zeppelin.gg/docs) | — | — | — | **no anti-nuke as a shipped product** | | [Beemo](https://beemo.gg/subscribe) | — | — | — | **join-rate only, no anti-nuke** | ## Wick — anti-nuke as the whole product, and the only repair attempt Wick watches the widest slice of the nuke sequence of anything in this comparison: mass deletion, dangerous permission grants to **any** role, and vanity URL changes all trigger a quarantine on the actor automatically. That last detail matters because it means the protection defends its own configuration surface rather than leaving it open — most products in this category guard the server and leave the door to their own settings unlocked. Wick is also the only bot found anywhere in this research that attempts **recovery** rather than only detection and punishment. Panic Mode deploys an isolated process to reconstruct pre-nuke state, then restores a structure-only backup image. Every other product here stops at "it happened and we logged it"; Wick tries to put the server back. **The caveat that matters, and it is not small.** Wick's anti-nuke explicitly exempts anyone holding a role above Wick's own in the hierarchy. An insider who can move roles — a compromised co-owner, for instance — gets past it entirely, and nothing in Wick's documentation indicates that edit itself gets recorded anywhere. Pricing is $5/month per server, or $20/month covering 12 servers, about $1.67 per server for anyone running several communities. ## Security Bot — free anti-nuke, paid automod Security Bot inverts the category's usual shape: **anti-nuke and anti-raid are free**, and chat automod is the premium feature — the opposite of almost everything else in this research. A free Security Bot server is protected from its own administrators and not from ordinary spam, which is worth knowing before assuming the usual free-tier shape applies. Its exemption rule mirrors Wick's: anyone at or above the bot's own role escapes detection. Pricing shows $3.49/month in bundles of 1, 3 and 7 servers on its pricing page, while the documentation states a different figure — check the checkout directly before committing. ## ProBot — anti-nuke included, anti-raid behind the paywall ProBot's Advanced Protection module gives it a genuine anti-nuke capability, putting it in the same small group as Wick and Security Bot. Its anti-raid — join-rate flooding specifically — sits behind payment; Advanced Protection itself is described on both of ProBot's pricing tiers, so check the tier descriptions directly rather than assuming from a summary. Tier 1 is $5/month or $40/year. ## Vetox — free protection across the widest rule count, paid recovery Vetox's approach to anti-nuke is a 20-rule protection system, free on Starter, covering mass channel and role deletion, permission changes, mass bans and kicks, and webhook and emoji abuse — the widest single rule count in this comparison. Every punishment type carries an **undo path**, a whitelist, per-role overrides and command limits, all configured from one dashboard rather than chat commands. Where Wick answers "what happens after a nuke" with server reconstruction, Vetox answers it with **Server Backup** at Max ($7.99/month): scheduled, cross-server-restorable snapshots of channels, roles, settings, bans, nicknames and role assignments. ## The bots that look like they should have this, and don't Four credible, deep moderation products in this research have **no anti-nuke at all**, checked against their own documentation and command lists. **Sapphire** is generous everywhere else — seven free Join Guard filters, a 12-module automod engine, 80+ log types — but explicitly lacks both join-rate anti-raid and anti-nuke. **Gaius** gives away a heat-and-strike engine and account-age screening with no anti-nuke layer on top. **Zeppelin**, an automod engine that is its whole specialism, has no anti-nuke as a shipped product, though one is technically constructible from its automod triggers feeding a counter — not the same as having one. **Beemo** is purely a join-rate raid detector and states plainly it has nothing else. None of the four is a weaker product for lacking this; they simply solve a different problem. ## Which one to install | If your threat model is | Install | | ------------------------------------------------------------ | -------------------------------------------------- | | A rogue admin, and you want the deepest coverage available | **Wick** | | The same threat, on a tighter budget | **Security Bot** | | You already run ProBot for its free leveling | **ProBot's Advanced Protection** | | You want free protection plus a real backup-and-restore path | **Vetox** | | Deep moderation but no nuke risk in practice | **Sapphire**, free, with nothing to configure here | ## Frequently asked questions **What exactly counts as a "nuke"?** Mass channel or role deletion, mass bans or kicks, dangerous permission grants to a role, and webhook or emoji abuse used to spam or redirect the community elsewhere — usually several of these in sequence, carried out by an account that already had legitimate access. **Can anti-nuke protection be turned off by an attacker?** Often, yes, if that attacker outranks the bot. Wick's anti-nuke exempts anyone holding a role above Wick's own; Security Bot's exempts anyone at or above its own role. Keeping the bot's role at the top of the hierarchy and restricting who can move roles closes most of that gap. **Does anti-nuke protection undo the damage, or only stop it from continuing?** Usually only the latter. Wick's Panic Mode is the only response in this research that attempts to reconstruct a server after an attack. Vetox answers the same question with a separate, paid backup system rather than an automatic in-attack restore. **Is anti-nuke free anywhere?** Yes — Security Bot gives it away entirely, and Vetox's 20-rule protection system is free on Starter. Wick and ProBot both gate their anti-nuke behind a paid tier. ## Method Every capability and price above was read from the vendor's own site or documentation on 31 August 2026. The "exactly four" figure is scoped to the moderation and security bots examined across this research, not to Discord's full security category.