# facebookresearch/segment-anything

The repository provides code for running inference with the SegmentAnything Model (SAM), links for downloading the trained model checkpoints, and example notebooks that show how to use the model.

Repository: https://github.com/facebookresearch/segment-anything
Canonical: https://ross.abutalabs.com/products/segment-anything
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Last push: 2024-09-18T17:46:55+00:00

## Health v2 (maintenance only)
Score: 30/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 89
- inputs: {"age_days": 1259, "days_push": 714, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 54759, forks 6355 (observed 2026-08-28T04:12:18.324463+00:00)

## What it is
Segment Anything Model (SAM) from Meta AI is a promptable image segmentation foundation model that produces high-quality object masks from point or box prompts, with strong zero-shot performance. This repository provides inference code, pretrained model checkpoints, and example notebooks for using the model.

## Use cases
- segment all objects in an image
- generate object masks from point or box prompts
- zero-shot image segmentation
- extract masks for image editing pipelines
- label segmentation data for training other models
- run SAM inference in a Python notebook

## When to choose
- you need high-quality promptable segmentation masks without training a custom model
- you want a well-tested, widely adopted segmentation foundation model with pretrained checkpoints
- you need zero-shot segmentation across diverse visual domains

## When to avoid
- you need video segmentation or streaming memory - use SAM 2 instead
- you need real-time inference on CPU-constrained hardware
- you need semantic class labels rather than class-agnostic masks

## Facets
- artifact type: library
- maturity: stable
- function: machine-learning, computer-vision, image-processing, deep-learning
- domain: computer-vision, image-processing, artificial-intelligence, machine-learning
- platform: python, cross-platform
- tags: segment-anything, image-segmentation, foundation-model, promptable-segmentation, object-masks, zero-shot, pytorch, meta-ai, gpu

## Member repositories
- facebookresearch/segment-anything (main) score 30

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:12:18.324463+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:19:08.016924+00:00, confidence not recorded.
  - readme: https://github.com/facebookresearch/segment-anything (fetched 2026-08-28T04:12:18.324463+00:00, sha 942cfbb82784)
- Data as of 2026-08-30T08:39:29.467469+00:00.