Introducing Meta Segment Anything Model 3 (SAM 3)

With SAM 3 you can use text and visual prompts to precisely identify, segment, and follow any object in images or videos