SAM (Segment Anything Model) by Meta

A promptable, open-source AI model designed for real-time object segmentation in images and videos.

• Supports segmentation via points, boxes, masks, or text prompts
• Trained on the SA-1B dataset with over 1.1 billion masks across 11 million images
• Achieves strong zero-shot performance without task-specific fine-tuning
• Applicable in image editing, robotics, AR/VR, medical imaging, and autonomous systems
• Open-source under Apache 2.0 license

CLICK HERE TO DISCOVER SAM