A Windows desktop application for efficient image captioning using OpenAI, Anthropic, or local models via Ollama and LM Studio.
tagmeister helps you organize and caption your datasets with ease.
- Browse and view images from any directory
- Generate AI-powered image captions using GPT models (OpenAI, Anthropic, Ollama, LM Studio)
- Choose between cloud (OpenAI, Anthropic) and local (Ollama, LM Studio) captioning backends
- Edit captions manually, save captions automatically
- Keyboard shortcuts for efficient navigation
- Batch processing support
- Resizable panels for customizable layout
- Windows 10 or later
- For cloud captioning: OpenAI or Anthropic API key
- For local captioning: Ollama and/or LM Studio installed (if using local models)
- CORS must be enabled on your LM Studio server and Ollama for local models to work.
How to enable CORS for Ollama
- Download the latest release (
tagmeister-portable.exe
) - Run the executable (no installation required)
- Enter your API key(s) in the settings (OpenAI and/or Anthropic, if using cloud models)
- (Optional) Install Ollama or LM Studio for local model support
Important: Make sure CORS is enabled on your LM Studio server and Ollama.
How to enable CORS for Ollama - Start captioning your images!
# Clone the repository
git clone https://github.com/oshtz/tagmeister-pc.git
# Navigate to the project directory
cd tagmeister
# Install dependencies
npm install
# Run the app in development mode
npm run tauri dev
To build a release version for Windows:
npm run tauri build
- Click the folder icon to select an image directory
- Select images from the left panel
- Choose your preferred captioning backend in the settings (OpenAI, Anthropic, Ollama, or LM Studio)
- Click "Generate" to create AI captions
- Edit captions as needed
- Captions are automatically saved as
.txt
files alongside your images
- API keys are stored locally in app preferences
- No data is sent to external servers except to the selected AI backend (OpenAI, Anthropic, or your local Ollama/LM Studio instance)
- All image processing is done locally except for caption generation via cloud APIs
Contributions are welcome! Please feel free to submit a Pull Request.