Imaginator is a lightweight, local desktop application for AI image generation. Powered by modern C++20/23 and a fluid Qt6 QML chat interface, it runs models like Stable Diffusion, FLUX, etc., natively without Python or Docker overhead. Fast, private, fully cross-GPU compatible, with an automatic CPU fallback
Imaginator provides an instant-start, low-memory footprint desktop experience powered directly by C++.
- Conversational UI (Chat-like): A seamless, dark-mode-first QML interface. Type a prompt, and the generated image appears in the chat thread (Txt2Img & Img2Img).
- Cross-Vendor GPU Acceleration: Powered by
stable-diffusion.cppwith a primary focus on Vulkan, NVIDIA, AMD, Intel and a seamless CPU fallback for testing.. - Smart Model Management: Easily load
.safetensorsfiles locally. The app automatically detects the model architecture (SDXL, FLUX, etc.) and dynamically adjusts the UI parameters. - Strict Modular Architecture: Complete separation between the pure C++ Core (inference, hardware detection) and the Qt UI layer, following SOLID principles.
Capture Moment is built for performance and modularity, using best-in-class libraries for each technical challenge.
| Component | Technology | Role |
|---|---|---|
| Language | C++20 / C++23 | Core Logic, Speed, and Control. |
| Build System | CMake 4.2+ | Cross-platform build configuration. |
| AI Backend | stable-diffusion.cpp | Pure C/C++ inference engine (SD, FLUX, Qwen-Image). |
| UI | Qt6 Quick/QML | Modern, declarative, and hardware-accelerated user interface. |
Imaginator supports Windows, macOS, and Linux. Due to the use of modern C++23 and Vulkan, specific SDKs are required.
For detailed, step-by-step instructions on setting up the Vulkan SDK, Qt, and compiling the project