Tag
SAM 3.1 can detect, segment, and track objects in images and video using text prompts on the Meta Model API, with specified inference costs.