SAM3 Demo

SAM3 Demo

Try the SAM3 segmentation model in a Gradio app: point it at a picture with a prompt and see the objects it picks out. Runs in your browser or on your own computer.

Made by prithivMLmods on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.

About

What it is: This Space is a Gradio demo of Meta's SAM3 model, a segmentation model you steer with prompts. Segmentation means marking which pixels of a picture belong to which object.

The page lets you try the model in a browser without setting it up yourself.

Who it suits: It suits people who want to see what SAM3 can pick out of an image, and developers who want a short Gradio example that loads one model.

Industry | Who uses it | How they use it | Example
Computer vision | Researchers | Try prompts on test images | Checking how well SAM3 separates one object from the rest
Design | Image editors | Look at the masks the model produces | Judging whether it could help cut out a subject
AI development | Developers | Read app.py and requirements.txt | Reusing the app for another segmentation model
Education | Students | Open the page and try the examples | Learning what promptable segmentation means

Skip it if you need to process a whole folder of images, since this is a single demo page.

Where to use it: The app runs in these places.

• A web browser, on the Space's page at Hugging Face
• A local copy on Windows, macOS or Linux, running Python
• Hugging Face's zero-a10g hardware, where the online Space runs

Use case | Who | What they get back
Try SAM3 online | Anyone curious | A result in the browser page
Test prompts on an image | Researchers | The objects the model picked out
Study the demo code | Developers | A short Gradio app around one model

How to set it up: Online you only need a browser. A local copy needs these.

• Python on Windows, macOS or Linux
• The pip command
• Room on disk for the SAM3 model

1. Install the huggingface_hub package.
2. Download the Space's files into a folder with the hf command.
3. Move into that folder.
4. Install gradio at version 6.30.0.
5. Install the packages listed in requirements.txt.
6. Start the app with python app.py.

pip install -U huggingface_hub
hf download prithivMLmods/SAM3-Demo --repo-type space --local-dir SAM3-Demo
cd SAM3-Demo
pip install gradio==6.30.0
pip install -r requirements.txt
python app.py

How to use it: Open the Space's page, or run the command below in your copy and open the page the app serves. Then give it an image and a prompt and run it.

python app.py

Example: You open the Space, add a picture, type a prompt naming the object you care about and run it. The page shows the picture with that object picked out.

For newcomers and experts: If you are new, open the Space's page and start with one of the example pictures. Change the prompt and watch what the model selects.

If you know the field, the app wraps a single model, facebook/sam3, in Gradio 6.30.0. Online it runs on zero-a10g hardware, and a local copy downloads the model weights on top of the Python packages. The repository is short, with app.py, an examples folder and two requirement files.

Open the Space page or run app.py
Add an image and a prompt
SAM3 segments what the prompt names
The result shows on the same page

Good to know: The repository has 10 files in all. Hugging Face lists the Space with Gradio and MCP tags.

• License: apache-2
• Cost to run: none named
• Data it reaches: only the model repository it names. The source says nothing else.

On Hugging Face, 13 November 2026

• Likes: 91
• Built with: gradio
• License: apache-2.0

Links

• Space: https://huggingface.co/spaces/prithivMLmods/SAM3-Demo
• Files: https://huggingface.co/spaces/prithivMLmods/SAM3-Demo/tree/main

Try it

Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.