According to @DeepLearningAI, Meta released an open-weights computer vision suite featuring SAM 3, SAM 3D Objects, and SAM 3D Body. According to @DeepLearningAI, SAM 3 segments images and video and supports text-prompted segmentation. According to @DeepLearningAI, SAM 3D Objects converts segmented items into 3D meshes or Gaussians and can optionally use point clouds. According to @DeepLearningAI, SAM 3D Body generates full 3D human figures. According to @DeepLearningAI, Meta reports the models outperform most rivals on segmentation and 3D quality in its internal tests. According to @DeepLearningAI, all models are available to try online with downloadable weights under the Meta license. According to @DeepLearningAI, the announcement was shared on December 8, 2025, and the source does not mention any cryptocurrency integrations or token impacts.

According to AI at Meta, SAM 3 is a unified model that enables detection, segmentation, and tracking of objects across images and videos, source: AI at Meta (X post, Nov 19, 2025). AI at Meta states SAM 3 introduces text and exemplar prompts to segment all objects of a target category, source: AI at Meta (X post, Nov 19, 2025). The announcement comes via Meta’s official AI account with no details provided on release timing, licensing, datasets, or code availability, source: AI at Meta (X post, Nov 19, 2025). For traders, this is a product capability update from Meta’s AI group focused on video-capable computer vision and category-wide segmentation; the post includes no crypto or blockchain references, so any crypto-market impact would be indirect, source: AI at Meta (X post, Nov 19, 2025).