
Aimmy is a source-available external tool that performs AI-based aim alignment through screen capture, ONNX, and YOLOv8 detection. It runs outside the game client rather than injecting code, reading the display and adjusting aim based on detected targets. Originally built by Babyhamsta and contributors including MarsQQ and Taylor, it is positioned as a "second eye" and accessibility-oriented aim assist rather than a Roblox script executor.

The tool uses YOLOv8 target detection loaded through ONNX models, with DirectML acceleration for GPU-based inference. Main functions include:
Because detection is visual, the same build applies across multiple games instead of a single title.
Aimmy targets Windows PC only and requires the .NET 8 runtime, since the interface is built with C# and WPF. It supports both NVIDIA and AMD GPUs through DirectML. The screen-capture approach means it is not a mobile tool and does not attach to processes, so there is no auto attach or injection step. Reported use spans Roblox and external titles such as Fortnite, Apex Legends, Battlebit Remastered, MW2, and MW3, with per-game configs and models adjusting confidence and field of view. Results depend on the model matched to each game.
Distribution and release activity are tracked on GitHub, with additional builds referenced on SourceForge and discussion across Discord and community boards. Aimmy became officially source-available with its v2 line, and version history is visible through public release pages. When a build stops opening, common causes are a missing .NET 8 runtime, an outdated ONNX model, or GPU driver issues. Reviewing release notes and matching the latest version to your hardware resolves most launch and detection problems.
Aimmy is free with no key system, so access does not depend on keyless unlock steps or paywalls. Practical highlights:
The tool functions as an external AI aim utility, not a code injector or script hub. Before use, confirm .NET 8 is installed, verify GPU compatibility, and load a model that matches your target game for accurate detection.