System Requirements
Minimum Requirements
Tag-AI can run on relatively modest hardware, but performance will be affected.
- Operating System: Windows 10/11, macOS 10.15+, or Linux (Ubuntu 20.04, Fedora 34, or
equivalent)
- Processor: Quad-core CPU @ 2.0 GHz or better
- Memory: 8GB RAM minimum
- Storage: 5GB free space for application and AI models
- Internet: Required for initial setup and Gemini API (if used)
Recommended Specifications
For optimal performance, especially when processing large image libraries:
- Processor: 8-core CPU @ 3.0 GHz or better
- Memory: 16GB RAM or more
- Storage: SSD with 10GB+ free space
- GPU: NVIDIA GPU with CUDA support or AMD GPU with ROCm support
GPU Requirements
GPU acceleration significantly improves processing speed:
NVIDIA
- CUDA 11.8 or newer (12.4 recommended)
- Minimum: NVIDIA GeForce GTX 1060 6GB or equivalent
- Recommended: NVIDIA RTX 2060 or better
AMD
Mac
- Apple Silicon (M1/M2/M3) or Intel Mac with discrete GPU
- macOS uses Metal API automatically, no additional setup required
Tag-AI can run without a GPU, but processing speed will be significantly reduced.
Disk Space
Tag-AI requires the following disk space:
| Component |
Space Required |
| Application |
~250MB |
| Ollama |
~100MB |
| LLaVA Model |
~4GB |
| Database (varies by library size) |
~20MB per 1000 images |
| Total Minimum |
~5GB |
Standard Formats
- JPEG (.jpg, .jpeg)
- PNG (.png)
- GIF (.gif)
- TIFF (.tif, .tiff)
- BMP (.bmp)
- HEIF/HEIC (.heif, .heic)
RAW Formats
- Nikon (.nef)
- Canon (.cr2, .cr3)
- Sony (.arw)
- Fujifilm (.raf)
- DNG (.dng)
- Other RAW formats: .rw2, .orf, .srw, .pef, .mos
OS-Specific Requirements
Windows
- Windows 10 (20H2 or later) or Windows 11
- Microsoft Visual C++ Redistributable 2019 or newer
- Administrator privileges for installation
macOS
- macOS 10.15 (Catalina) or newer
- Apple Silicon (M1/M2/M3) or Intel processor
Linux
- Ubuntu 20.04 or newer, Fedora 34 or newer, or equivalent distribution
- Qt 6 libraries (libqt6-*)
- Python 3.9 or newer