Caption Creator is a app that helps you create text for images. It can write natural captions or clean comma-separated tags, so it is useful for image datasets, LoRA training, AI image prompting, or organizing image folders.

The app runs on your own computer. Your images stay local, and you can use either the built-in local models or your own model through LM Studio.

Official Website: caption-creator.merserk.com

Online version: aitools.merserk.com/caption-creator

Support the developer on Patreon: patreon.com/MM744

Features:

• Captions or Tags: Choose between full image descriptions or short comma-separated tags.

• Clean Tag Output: Tags are automatically cleaned and formatted into one easy-to-copy line.

• Single Image or Batch Mode: Work with one image, or process many images in one run.

• Portable App: No full setup needed. Unpack the folder and run the app.

• Local AI Option: Use included local models so images are processed on your own computer.

• Modern Dark Interface: Clear dark UI with live status, progress, and preview panels.

• LM Studio Support: Use a model already running in LM Studio instead of the built-in model option.

• Paste Images Directly: Paste an image from your clipboard in Single Image mode with Ctrl+V.

• Model Management: Download or delete the built-in models from inside the app.

• Prompt Enrichment: Add extra instructions when you want the AI to focus on something specific.

• Save as ZIP: Package the current result folder into a ZIP file with one click.

• Model Choices by GPU Memory: Pick from 6GB, 8GB, or 10GB+ model options depending on your graphics card.

• Low-VRAM Mode: Helps the app run on lower-memory graphics cards, but generation may be slower.

• Keep Model Loaded: Keep the model ready after a run so the next generation can start faster.

• Flexible Text Controls: Add trigger words and choose the maximum caption or tag length.

• Easy Result Actions: Copy the generated text or open the output folder from the app.

How to Use:

1. Download and Unpack
   Download the program and unpack the .zip archive into a folder.

2. Launch the Application
   Double-click Caption Creator.exe to launch the program.

3. Manage Your Model

   • Click the "Model / VRAM Configuration" button to open the model list.

   • To use a built-in model: Click Download next to the model that matches your graphics card.

   • To use LM Studio: Open LM Studio, start its local server, load a vision model, then select Custom (LM Studio) in Caption Creator.

   • Select the model you want to use.

4. Load Image(s)

   • Single Mode: Drag & drop an image, click to browse, or paste an image from your clipboard.

   • Batch Mode: Drag & drop multiple images or click to select several images.

5. Configure and Generate

   • Choose Captions or Tags.

   • Adjust options such as Max Words, Trigger Words, Prompt Enrichment, Low-VRAM Mode, Keep Model Loaded, or Shut down PC after generation.

   • Click Generate.

6. Get Results
   Processed images and matching .txt files are saved in the output folder, organized by mode, date, and run number.