FiveM Super Jump Detection: Catching Vertical Velocity Cheats

Super-jump cheats are deceptively simple: they either amplify the initial jump impulse beyond what GTA V's physics engine allows or set gravity to near zero so the player floats indefinitely. Either way, they break roleplay economies, let cheaters reach unreachable loot spots, and make combat completely one-sided. Catching them requires understanding exactly what the physics engine permits, and checking every player against that boundary on every tick, server-side.

What Super Jump & Gravity Cheats Actually Do

GTA V's physics engine caps the vertical launch velocity a standard jump can produce. On foot, that ceiling sits at roughly 4.5 m/s upward at the moment of takeoff, with gravity pulling the player back down at a known, fixed rate. A super-jump cheat circumvents this in one of two ways.

The first approach inflates the launch impulse directly, often by patching the in-memory jump force multiplier or by calling a native like SET_SUPER_JUMP_ENABLED outside of any valid server-side activation. The player rockets upward at 15, 30, or even 100 m/s, a value that is simply impossible under normal conditions.

The second approach zeroes out or inverts gravity. The player jumps normally but never comes back down, drifting across the map until the cheat is toggled off. The vertical velocity stays near zero across dozens of ticks while the player remains airborne, which is equally detectable once you know what the fall curve should look like.

Both signatures are distinct from noclip, where horizontal displacement is the primary tell. Super-jump leaves the player fully subject to horizontal physics but breaks the vertical channel specifically.

Why Vertical-Velocity Thresholds Are the Right Layer

The naive approach is to watch maximum airborne height. The problem: a player who jumps off a skyscraper rooftop or falls from a helicopter legitimately reaches enormous heights. Height alone cannot distinguish cheating from normal play on a map as vertical as Los Santos.

Velocity at takeoff is much harder to fake convincingly. GTA V reports player velocity to the server on every sync frame. When that Z-component spikes past the physics ceiling in a single tick and the player was grounded the previous tick, there is no legitimate explanation. The cheat injected an impulse the engine would never generate on its own.

The same principle applies to the gravity-off variant, just inverted. Rather than a spike, the detector looks for a flat or near-zero Z-velocity that persists while the player is flagged as airborne across successive ticks. Physics says velocity should be increasing negatively (the player accelerating downward). When it isn't, gravity has been disabled.

This is the same family of thinking described in our article on speed hack detection: compare reported state against what the physics engine could physically produce, and treat any divergence as a violation signal rather than treating max-height or max-distance as the primary metric.

The Detection Pipeline: Tick by Tick

A robust super-jump detector runs the following checks on every position-sync event from a player's client agent:

• 1. Takeoff spike check. If the player was grounded on the previous tick and the current Z-velocity exceeds the engine's maximum launch velocity (with a small tolerance for ping-induced jitter), a violation counter increments.
• 2. Hang-time analysis. While the player is airborne, the expected Z-velocity is calculated from the known gravity constant and elapsed time since takeoff. If the reported velocity deviates from the expected fall curve by more than the jitter margin across three or more consecutive ticks, a gravity-off flag is raised.
• 3. Ping compensation. Each player's rolling average ping is used to widen the jitter tolerance proportionally. A 200 ms player gets a slightly wider window than a 20 ms player. This prevents lag spikes from reaching enforcement thresholds.
• 4. Multi-tick confirmation. A single anomalous tick never triggers enforcement. The system requires a configurable number of consecutive or near-consecutive violations before issuing a kick or ban, eliminating the false positives that plague simpler threshold-only systems.
• 5. Context cross-reference. The violation is cross-checked against recent events: was a vehicle involved, was the player on a slope, did a script legitimately grant super-jump? Whitelisted server-side events suppress the detector for their duration.

Interaction With Behavioral Signals

Super-jump is rarely the only cheat running. Players who use gravity cheats frequently combine them with godmode, weapons spawning, or economy exploits during the same session. A standalone velocity check catches the jump itself, but the broader picture matters for banning with confidence rather than kicking and hoping the player doesn't reconnect.

This is where behavioral detection complements individual module checks. When a super-jump violation coincides with anomalous combat accuracy, inventory mutations, or event-spam patterns in the same session window, the confidence score climbs quickly. A high-confidence composite signal means the ban is defensible in review: you have multiple independent data streams all pointing at the same player within the same timeframe.

Fiveuxe's SENTINEL AI security co-pilot sits above the 27 individual detection modules and correlates exactly this kind of cross-signal evidence. A super-jump flag on its own might be held for human review on a borderline ping reading. The same flag paired with weapon-spawn events and godmode indicators from the same session gets actioned automatically.

Server-Side Only: No Kernel Driver Required

Every check described above is decided on the server. Fiveuxe deliberately avoids kernel drivers: there is no ring-0 component, and no separate process reading another process's memory. The sub-millisecond client agent reports position and event data to the server, and all analysis happens there.

This matters for two reasons. First, it means the detection cannot be bypassed by kernel-level rootkits that hide cheats from client-side scanners, because the server is not scanning the client at all: it's checking physics consistency, which the cheat cannot hide without also preventing itself from working. Second, players add zero FPS cost: the agent is lightweight by design, and all the heavy comparison logic runs server-side where the cheater has no visibility into it.

When a super-jump ban fires, the evidence package, including the flagged velocity readings, the tick timeline, and any corroborating signals, is posted to the web panel and sent as a real-time Discord webhook. Staff reviewing the ban can see the exact readings that triggered enforcement rather than a vague "cheating detected" notice.

Bans feed into Fiveuxe's Hive-Mind global ban network. A player banned for super-jump on one server is rejected across 100+ connected servers in approximately 0.3 seconds. Shadow-Link identity enforcement tracks the ban across HWID, Discord IDs, IP, and 12 other identifiers, so creating a new Steam account doesn't get the cheater back in.

FAQ