Levelmodel
Classvision
Parameters848M
Contextnot in the notes
ReleasedNov 2025
Opennessopen weights restricted
Runs onsingle datacentre GPU
Statusactive
Benchmark Results
vendor-reported
LVISdomain
instance segmentationvendor-reported
48.5
02550
Score in AP. Higher is better.
boxvendor-reported
40.6
03060
Score in cgF1. Higher is better.
instance segmentationvendor-reported
37.2
03060
Score in cgF1. Higher is better.
SA-Co/Golddomain
human 72.8vendor-reported
54.1
03060
Score in cgF1. Higher is better.
SA-Vdomain
test; human 53.1vendor-reported
30.3
03060
Score in cgF1. Higher is better.
Where this comes from
Every score names who ran it: vendor-reported, maintainer-run or independent. A rank is shown with the population it was ranked in (gap rule 3).
Lineage
successorfinetune
SAM 2 is the parent (successor)high confidencemodel cardSAM 3.1 (Object Multiplex) is a child (successor)high confidencemodel cardSAM 3D Objects is a child (finetune)low confidencenotesWhere this comes from
Parent and child relations from model cards, Hugging Face base_model fields and the genealogy tables of the research notes. The tree shows two generations each way, 5 models. Confidence and evidence are the notes' own.
R&D Topics
Identity
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.
Architecture and Size
- Architecture
- dense transformer Shared vision encoder; DETR-based detector conditioned on text, geometry and image exemplars; tracker inherits SAM 2 encoder-decoder; presence token
- Total parameters
- 848M sources
- Active parameters
- –
- Checkpoint parameters
- 859.9M
- Note
- 848M per repo; HF facebook/sam3 safetensors 859.9M
- Context length
- not in the notes
- Input
- image, video, text
- Output
- image, video
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).
Openness and Access
- Openness
- open weights restricted
- Licence
- SAM License sources
- Access routes
- open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.
Deployment
- Runs on
- single datacentre GPU
- Footprint
- Single GPU (judgement)
- Runtimes
- not in the notes
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.
Indonesia
- Indonesian support
- not stated
- Regional languages
- not in the notes
- Jakarta hosting
- unknown
- Languages
- not in the notes
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.
Industry Use Cases
No industry use case links this model.
Links
DeveloperMetabig tech, USA
Metabig tech, USAModelSAM 3Meta, vision
SAM 3Meta, visionR&D TopicsPromptable Segmentation Foundation Models (Segment Anything)Open-Vocabulary Object Detection and Grounding
Use CasesMedical Image SegmentationHealthcareUse cases of models in the same R&D topic; none is recorded for this model itself.
Who Runs Itnone recordedDeployments of these use cases; none names this model.
Models 3
Organisations 1
R&D topics 2
Open-Vocabulary Object Detection and GroundingDetectors and segmenters that find objects named in free text, including classes never la…produced byPromptable Segmentation Foundation Models (Segment Anything)Models that cut out any object from a point, box or text prompt, trained on very large ma…exemplar
Sources
- SAM 3 Introduces a More Capable Segmentation Architecture for Modern Vision Workflows infoq.com https://www.infoq.com/news/2025/11/meta-sam3
- GitHub - facebookresearch/sam3: The repository provides code for running inference and finetuning with the Meta Segment Anything Model 3 (SA github.com https://github.com/facebookresearch/sam3
- huggingface.co/api/models/facebook/sam3 https://huggingface.co/api/models/facebook/sam3
To cite this page: Project AURION, AI Use-Cases Research & Innovation, "SAM 3", https://aurion.semai.id/models/meta-sam-3/, records of October 2026.