Meta Launches SAM 3.1 for Text-Prompted Image and Video Segmentation
September 21, 2026
The Segment Anything Model (SAM) 3.1 enables object detection, segmentation, and tracking in images and video using text prompts via the Meta Model API. Usage is billed on a per-image or per-video-frame basis.
HOW THIS AFFECTS YOU
●
builderYou can integrate precise, promptable segmentation into video editing or computer vision pipelines via API.
●
designerYou can use text-based prompts to automate complex object isolation in creative workflows.