We're a charity on a mission to educate and certify the upcoming wave of developers in the world of Computer Engineering 🌍. Need assistance? Hop into our Discord and toss your questions at @Wonder in the #ai-aimbot channel (be sure to stick to this channel or face the consequences! 😬). Type away your query and include @Wonder in there.

Our AI Aimbot 🤖 sharpshoots targets in any game with humanoid characters, harnessing the power of YOLOv5. Currently, it's a ninja against anti-cheat systems, as it's visual-only. Still, watch out for manual player reports! 👀

Intended for educational use 🎓, our aim is to highlight the vulnerability of game devs to AI-driven cheats. Pass it along to your game developer buddies, and save their games from being outsmarted!

⚠ Use at your own risk! If you're caught... well, you've been warned!

📹 Instructional Media

There are 3 Versions 🚀🚦🖥️

Fast 🏃‍♂️ - main.py ✅ Easy to set up, Works on any computer 💻
Faster 🏃‍♂️💨 - main_onnx.py ⚙️ May need to edit a file, Works on any computer 💻
Fastest 🚀 - main_tensorrt.py 🏢 Enterprise level hard, Works on computers with Nvidia GPUs only 🎮

🧰 Requirements

Nvidia RTX 980 🆙, higher or equivalent
And one of the following:
- Nvidia CUDA Toolkit 11.8 DOWNLOAD HERE

🚀 Pre-setup Steps

Download and Unzip the AI Aimbot and stash the folder somewhere handy 🗂️.
Ensure you've got Python installed (like a pet python 🐍) – grab version 3.11 HERE.
- 🛑 Facing a python is not recognized... error? WATCH THIS!
- 🛑 Is it a pip is not recognized... error? WATCH THIS!
Fire up PowerShell or Command Prompt on Windows 🔍.
To install PyTorch, select the appropriate command based on your GPU.
- Nvidia pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
- AMD or CPU pip install torch torchvision torchaudio
📦 Run the command below to install the required Open Source packages:

pip install -r requirements.txt

🔌 How to Run (Fast 🏃‍♂️ Version)

Follow these steps after Python and all packages have been installed:

Open PowerShell ⚡ or Command Prompt 💻.
Input cd , then drag & drop the folder containing the bot code into the terminal.
Hit Enter ↩️.
Type python main.py and press Enter.
Use CAPS_LOCK to toggle the aimbot 🎯. It begins in the off state.
Pressing q 💣 at ANY TIME will shut down the program.

🔌 How to Run (Faster 🏃‍♂️💨 Version)

Follow these steps after Python and all packages have been installed:

Open the config.py 📄 file and tweak the onnxChoice variable to correspond with your hardware specs:
- onnxChoice = 1 # CPU ONLY 🖥
- onnxChoice = 2 # AMD/NVIDIA ONLY 🎮
- onnxChoice = 3 # NVIDIA ONLY 🏎️

IF you have an NVIDIA set up, run the following

pip install onnxruntime-gpu
pip install cupy-cuda11x

Follow the same steps as for the Fast 🏃‍♂️ Version above except for step 4, you will run python main_onnx.py instead.

🔌 How to Run (Fastest 🚀 Version)

Follow these sparkly steps to get your TensorRT ready for action! 🛠️✨

Introduction 🎬 Watch the TensorRT section of the setup video 🎥 before you begin. It's loaded with useful tips!
Oops! Don't Forget the Environment 🌱 We forgot to mention adding environmental variable paths in the video. Make sure to do this part!
Get Support If You're Stumped 🤔 If you ever feel lost, you can always @Wonder your questions in our Discord 💬. Wonder is here to help!
Install Cupy Run the following pip install cupy-cuda11x
CUDNN Installation 🧩 Click to install CUDNN 📥. You'll need a Nvidia account to proceed. Don't worry it's free.
Get TensorRT 8.6 GA 🔽 Fetch TensorRT 8.6 GA 🛒.
Unzip and Relocate 📁➡️ Open the .zip TensorRT file and move all the folders/files to where the CUDA Toolkit is on your machine, usually at C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8.
Python TensorRT Installation 🎡 Once you have all the files copied over, you should have a folder at C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\python. If you do, good, then run the following command to install TensorRT in python.
```
pip install "C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\python\tensorrt-8.6.1-cp311-none-win_amd64.whl"
```
🚨 If the following steps didn't work, don't stress out! 😅 The labeling of the files corresponds with the Python version you have installed on your machine. We're not looking for the 'lean' or 'dispatch' versions. 🔍 Just locate the correct file and replace the path with your new one. 🔄 You've got this! 💪
Set Your Environmental Variables 🌎 Add these paths to your environment:
- C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\lib
- C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\libnvvp
- C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.8\bin
Download Pre-trained Models 🤖 You can use one of the .engine models we supply. But if it doesn't work, then you will need to re-export it. Grab the .pt file here for the model you want. We recommend yolov5s.py or yolov5m.py HERE 🔗.
Run the Export Script 🏃‍♂️💻 Time to execute export.py with the following command. Patience is key; it might look frozen, but it's just concentrating hard! Can take up to 20 mintues.

python .\export.py --weights ./yolov5s.pt --include engine --half --imgsz 320 320 --device 0

Note: You can pick a different YOLOv5 model size. TensorRT's power allows for larger models if desired!

If you've followed these steps, you should be all set with TensorRT! ⚙️🚀

⚙️ Configurable Settings

*Default settings are generally great for most scenarios. Check out the comments in the code for more insights. 🔍 The configuration settings are now located in the config.py file!
CAPS_LOCK is the default for flipping the switch on the autoaim superpower! ⚙️ 🎯

useMask - Set to True or False to turn on and off 🎭

maskWidth - The width of the mask to use. Only used when useMask is True 📐

maskHeight - The height of the mask to use. Only used when useMask is True 📐

aaQuitKey - The go-to key is q, but if it clashes with your game style, swap it out! ⌨️♻️

headshot_mode - Set to False if you're aiming to keep things less head-on and more centered. 🎯➡️👕

cpsDisplay - Toggle off with False if you prefer not to display the CPS in your command station. 💻🚫

visuals - Flip to True to witness the AI's vision! Great for sleuthing out any hiccups. 🕵️‍♂️✅

aaMovementAmp - The preset should be on point for 99% of players. Lower the digits for smoother targeting. Recommended doses: 0.5 - 2. ⚖️🕹️

confidence - Stick with the script here unless you're the expert. 🧐✨

screenShotHeight - Same as above, no need for changes unless you've got a specific vision. 📏🖼️

screenShotWidth - Keep it constant as is, unless you've got reasons to adjust. 📐🖼️

aaDetectionBox - Default's your best bet, change only if you've got the know-how. 📦✅

onnxChoice - Gear up for the right graphics card—Nvidia, AMD, or CPU power! 💻👾

centerOfScreen - Keep this switched on to stay in the game's heart. ❤️🖥️

📊 Current Stats

The bot's efficiency depends on your setup. We achieved 100-150 CPS with our test specs below 🚀.

- AMD Ryzen 7 2700
- 64 GB DDR4
- Nvidia RTX 3080

💡 Tip: Machine Learning can be tricky, so reboot if you keep hitting CUDA walls.

🤝 Community Based

We're all about collaboration. Your contributions can earn you credit and potential 💰!

Want to volunteer? Have video or program ideas? Tell us!

⚠️ Known Cheat-Detectable Games

Splitgate (reported by a Discord user 🕵️‍♂️), EQU8 detects win32 mouse movement library.

🚀 Custom Aimbots and Models

Show off your work or new models via Pull Requests in customScripts or customModels directories, respectively. Check out the example-user folder for guidance.

🌠 Future Ideas

Mask Player to avoid false positives

Happy Coding and Aiming! 🎉👾

README.md Unescape Escape

🎯 World's Best AI Aimbot 🎮

🙌 Welcome Aboard!