Choose your project type to get started
✓ Draw bounding boxes around objects
✓ Multi-class object detection
✓ Train/validation split support
✓ Standard YOLO format export
Perfect for training YOLOv5, YOLOv8, or any object detection model
✓ Bounding box + 4 corner keypoints
✓ Top-left, top-right, bottom-right, bottom-left
✓ Perfect for document/card detection
✓ YOLO keypoint format export
Ideal for training models that need precise corner detection
✓ AI-powered caption generation
✓ DreamBooth trigger token support
✓ Manual caption editing
✓ SD/LoRA training format export
Create datasets for Stable Diffusion, DreamBooth, and LoRA training
💡 Tip: You cannot change the project type after creation, so choose carefully!