OPEN-SOURCE AI AIMBOT

The open-source AI aimbot, without the build.

RootKit’s AI-Aimbot, AIMr and Aimmish proved YOLO beats colour detection. They also assume you can wrangle Python, CUDA and ONNX. NeyraX ships the finished version.

Undetected by design · Instant delivery · Trained for Fortnite (stylized characters)

150+
detections / sec
97%
enemy detection
0
memory reads
Fortnite
trained & tuned for

The open-source AI aimbot scene did the hard intellectual work. RootKit’s AI-Aimbot showed a YOLO model could pick humans out of any game. AIMr made it universal. Aimmish forked it and added hardware input. They are real projects, they are free, and the code is public — if you want to read exactly how an AI aimbot works, go read theirs.

What none of them are is a product. You install Python. You install the right CUDA and torch for your GPU. You find a model. You export it to ONNX or TensorRT for your card. You set up a virtual controller and a device cloak. Somewhere in there you discover your GPU is new enough that half the wheels do not exist yet. Then you start tuning.

NeyraX is that same idea with the assembly already done, aimed at one game. The model is trained on Fortnite. The engine builds its own TensorRT engine for your GPU on first run. The installer handles the driver stack. And when the model or the engine improves, your app updates itself — you do not go pull a new commit.

Universal model vs. one game done properly

A universal aimbot uses a model trained to find people in general. That is a genuine achievement, and it is also why it struggles in Fortnite: stylized characters, hundreds of skins, builds thrown up mid-fight, smoke and shield effects. The thing it is best at — realistic human shapes — is exactly what Fortnite does not have.

NeyraX’s model is trained on 100,000+ real Fortnite frames and retrained on new gameplay every release. Narrow beats general when the target looks nothing like the training set. That is the whole trade: NeyraX will never help you in CS2, and it is not trying to.

The part nobody warns you about: your GPU

Getting a YOLO model to actually run fast is where most builds die. Torch has to match your CUDA, CUDA has to match your driver, and a brand-new card can mean waiting on wheels that do not exist yet. Then you still need to export the model to TensorRT for your specific GPU, because raw PyTorch inference is too slow to feed an aim loop.

NeyraX does that build itself, once, on your machine, on first launch — about a minute — and caches the result. You do not pick a backend, you do not export anything, and you do not read a CUDA compatibility table.

What the open-source projects give you that we do not

The source. You can read every line, change anything, run it on any game, and pay nothing. If you enjoy that, or you play something other than Fortnite, those projects are the right answer and we would rather you used them than bought something that does not fit.

NeyraX is closed and Fortnite-only. What you get for that is a model that keeps improving without you, an engine that gets fixes pushed to it, and support when it breaks. Different deal, not a strictly better one.

Why players switch to NeyraX

Real neural detection

Detects enemies at 150+ FPS, any distance, any skin — retrained on real matches every update.

Zero-shake tracking

Velocity-gated stabilization kills detection jitter before it hits your crosshair.

Undetected by design

External only — no injection, no memory reads. The game sees controller input, nothing else.

Flick assist

Flick to a target and the engine finishes the last pixels. Overshoot-free.

Controller-native

Corrections blend into your own stick input. It amplifies your intent, not replaces it.

One-click installer

Sets up the AI engine, drivers, HidHide and your controller automatically. No terminals.

Pick your edge

Launch week: 25% off with code NEYRAX25

Day pass
$9.99/ 24h
Get Day pass
Weekly
$24.99/ week
Get Weekly
Most popular
Monthly
$39.99/ month
Get Monthly
Lifetime
$149.99one-time
Get Lifetime

Questions, answered

Is NeyraX built on RootKit’s AI-Aimbot or AIMr?+

No. NeyraX is its own engine with its own Fortnite-trained model and its own tracking. It shares the underlying idea — a YOLO-class network reading the screen — which is the same idea those projects popularised.

Why not just use the free one?+

If you can set up Python, CUDA and ONNX and you enjoy tuning, do. You will pay nothing. NeyraX is for people who want the Fortnite-trained model and the one-click setup, and who would rather not lose a weekend to a compatibility table.

Does NeyraX work on other games?+

No. The model is trained on Fortnite and only Fortnite. A universal open-source aimbot covers more games; it just does not know what a Fortnite skin looks like.

Do I need Python installed?+

No. Everything is bundled in the installer — engine, drivers, model. Nothing to set up by hand.