Notes:
- API Key: Saved in your browser local storage. Not sent anywhere except to Google.
- Sequential vs. Batch: The default mode processes all images in fast batches. Enable Process Sequentially to analyze each image one-by-one. This is slower and uses more API credits, but is necessary for a per-image delay.
- Generation Types: Choose 'Caption' for sentences, 'Tags' for keywords, or 'Weight Brief' for prompts formatted for AI image generation (e.g., `(word:1.2)`).
- Kohya_SS Export: When enabled, this creates a ZIP file with a folder structure and text files ready for training LoRA models.
- Auto Rename: Automatically renames image files and their corresponding text files to sequential numbers (e.g., 1.png, 1.txt) in the downloaded ZIP.
Options
Usage Instructions:
- Get your API key from Google AI Studio and paste it in the options.
- Click or drag-and-drop your images into the upload area.
- Select your desired Generation Type, Model, and Creativity level.
- Add optional Trigger Words or Prompt Enrichments.
- Enable any specific processing or export options (like Kohya_SS).
- Click "Start" to begin processing.
- Once complete, review the results and click "Download as ZIP".