metaai·lightalo unofficial · independent
Universe / Vision / SAM 3
Vision · 2025

SAM 3

Type 'yellow school bus' — it masks every single one

open source 848M params

Latest: SAM 3.1 (Mar 27, 2026) — multiplexing tracks up to 16 objects per forward pass, doubling video throughput from 16 to 32 fps on one H100

The first SAM that understands language: an 848M-parameter unified detector-plus-tracker that finds, segments, and tracks every instance of a concept from a short text phrase ('red baseball cap') or image exemplar, across images and video. Released November 19, 2025 with open checkpoints, alongside the browser-based Segment Anything Playground.

Why it matters

SAM 3 gave Segment Anything language: type 'yellow school bus' and it finds, masks and tracks every instance across images and video, reaching 75-80% of human performance on a 270,000-concept benchmark. It ships inside Instagram Edits and Meta AI, and the 3.1 update doubled video throughput to 32 fps on one H100.

Facts

Try it yourself

Lineage

Descends fromSAM 2
Led toSAM 3D (Objects + Body)SAM Audio

See the whole family tree →

Sources

More in Vision

DINOv3VGGTPerception Encoder & Perception Language ModelChameleonSapiensVideo Seal

Read the Vision story on the sky →

✦ Open on the map Explore Vision Quiz me