What is MLSAC? A machine-learning anti-cheat for Minecraft servers
MLSAC is a cloud anti-cheat for Minecraft Java servers that uses neural networks to detect KillAura, AimBot, AimAssist, TriggerBot, Reach and hits through walls.
MLSAC in one paragraph
MLSAC is an anti-cheat for Minecraft Java Edition servers built on machine learning. A lightweight plugin on the server watches how players fight — how they aim, click, move and hit — and sends that combat behaviour to MLSAC neural networks in the cloud. The models return how likely it is that the player is cheating, and the plugin turns that into alerts for staff and, if the owner wants, punishments or in-game suppression.
Unlike classic anti-cheats that check fixed rules (“was this hit too far?”), MLSAC learns what legitimate and cheating players look like from real recordings, so it can catch combat cheats that stay inside the usual rule limits.
What MLSAC detects
MLSAC focuses on combat cheats — the ones that decide PvP fights:
- KillAura — automatic aiming and hitting of nearby players
- AimBot and AimAssist — full or partial automatic aiming
- TriggerBot — automatic clicking when the crosshair is on a target
- Reach — hitting from further away than a normal client can
- WallHit — hitting players through blocks
Supported servers
The plugin runs on Paper, Spigot, Purpur and Folia, Minecraft 1.16.5 to 1.21, with builds for Java 17 and Java 21. Players can join from other client versions through ViaVersion.
Everything is configured on the website
Server owners manage MLSAC from the dashboard at mlsac.net. Settings are stored in presets: which worlds and WorldGuard regions are checked, which model runs, how sensitive detection is, who gets alerts, what happens on each flag, plugin messages in seven languages, staff menus and nametags above suspects. The plugin picks up changes by itself, without a restart.
Instead of an instant ban, MLSAC can suppress a suspected cheater: less damage dealt, more damage taken, lag, inventory tricks — first on rising detection buffer, then harder on each flag.
Models and plans
MLSAC runs several detection models in the cloud. Ultra is the newest model and catches the widest range of combat cheats; Pulse, Pro and Mini are the earlier models. There is a free plan, and paid plans raise the daily request limit and unlock the stronger models — see mlsac.net/pricing.
FAQ
- What does MLSAC stand for?
- MLSAC is a machine-learning (ML) anti-cheat for Minecraft servers. The project is also called MLS AntiCheat.
- Is MLSAC free?
- Yes, there is a free plan. Paid plans add more daily requests and stronger models.
- Does MLSAC lag the server?
- The heavy work — the neural networks — runs in the MLSAC cloud. The plugin only collects combat data and sends it, so the load on the server is small.
- Which cheats does MLSAC catch?
- Combat cheats: KillAura, AimBot, AimAssist, TriggerBot, Reach and hits through walls.