Model Tracker
SAM
SAM is an AI model made by Meta. Meta released it on 5 April 2023. It is generally available. Its weights are open. Its licence is Apache 2.0.
In inglese
- Released
- 5 April 2023 — ai.meta.com, read 16 October 2026
- Status
- Generally available — github.com, read 16 October 2026
- Takes
- Not confirmed: two readings disagree: text, image and image
- Returns
- Image — arxiv.org, read 16 October 2026
- Open weights
- Yes — github.com, read 16 October 2026
- Licence
- Apache 2.0 — github.com, read 16 October 2026
- Context window
- Not confirmed: The maker's cited pages do not state a context window in tokens for SAM.
- Longest answer
- Not confirmed: The maker's cited pages do not state a standard maximum output in tokens for SAM.
- Knowledge cutoff
- Not confirmed: The maker's cited pages do not state a knowledge cutoff for SAM.
- API price, input
- Not confirmed: The maker's cited pages provide no standard API input-token price for SAM.
- API price, output
- Not confirmed: The maker's cited pages provide no standard API output-token price for SAM.
- Where
- Downloadable weights — github.com, read 16 October 2026
- Parameters
- Not confirmed: The maker's cited pages describe multiple SAM backbone variants but do not publish a parameter count for a single standard model.
- Model card
- Not confirmed: two readings disagree: https://arxiv.org/html/2304.02643 and https://github.com/facebookresearch/segment-anything
Questions
When was SAM released?
Meta released SAM on 5 April 2023.
Is SAM still available?
Yes: it is generally available, as Meta states it.
Are the weights of SAM open?
Yes: Meta publishes the weights of SAM for download.
What is the licence of SAM?
The licence of SAM is Apache 2.0, as Meta names it.
More from Meta
- Audiobox
- AudioGen
- Chameleon
- CM3leon
- Code Llama
- DINOv2
- DINOv3
- Emu
- Emu Edit
- Emu Video
- EnCodec
- I-JEPA
- ImageBind
- LLaMA
- Llama 2
- Llama 3
- Llama 3.1
- Llama 3.2
- Llama 3.3
- Llama 4 Maverick
- Llama 4 Scout
- M2M-100
- Make-A-Video
- Movie Gen
Sources
- Introducing Segment Anything: Working toward the first foundation model for image segmentation (read 16 October 2026)
- facebookresearch/segment-anything (read 16 October 2026)
- Segment Anything (read 16 October 2026)