Lunar LITE is built on top of the original Lunar project.
It features an updated triggerbot, modernized packages, and YOLOv12 support.
Lunar V2 includes:
- ✅ 25+ customizable settings
- ✅ Built-in UI
- ✅ Improved detection accuracy
- ✅ Supports YOLOv8, YOLOv10, YOLOv12, and TensorRT
- ✅ Xbox controller support
- ✅ Logitech GHUB mouse input
- ✅ Makcu support (UDP)
- ✅ Works on AMD and NVIDIA graphics cards
Lunar uses screen capture + YOLO object detection to locate enemies in real-time.
It doesn’t touch memory or inject code — think of it as a robot that watches your screen and gives you precise X,Y coordinates of targets.
🎯 Preconfigured for Fortnite — some sensitivity tuning may be needed for other games.
Lunar LITE works with:
Finding good models has always been difficult (atleast for me). So we partnered with mossymodels to make an archive of models, and made a rating system so it's easy to figure out which models are actually good.
MossyModels is a growing community archive for YOLO models, built to make sharing and discovering models easier. Browse models by game, compare ratings and downloads, or upload your own ONNX/PyTorch (.pt) models for others to use.
📦 Installation
- Install Python 3.10.5
- Install CUDA Toolkit 11.8, 12.4, or 12.6 (12.6 recommended)
- Navigate to the root folder and run:
install_requirements.bat - Launch with:
start.bat
⚙️ Usage / Troubleshooting
- Make sure your installed CUDA version matches.
- Visit pytorch.org and install the right build.
Command for CUDA 12.6:
pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu126
python lunar.py
python lunar.py setup
python lunar.py collect_data
Support is only guaranteed for Lunar V2.
Please don’t expect full help for the free LITE version.
