Text, point and box prompts
Name an object in a Prompt, or mark it with Point Prompts or Box Prompts.
Find objects and outlines
Segment Anything Model 3
SAM 3 is a unified foundation model for promptable segmentation in images and videos. It can detect, segment, and track objects using text or visual prompts such as points, boxes, and masks.
Made by Meta50 creditsper run Pro
Opens the image studio with Segment Anything Model 3 ready to use. You can look around as a guest. Running it takes the Pro plan or higher.
What Segment Anything Model 3 makes, what it needs and what you get
What it makes
Find objects and outlines
Finds objects, depth, poses and outlines in a picture.
What you need
- The picture to work on
Optional: what to find.
What you get
Masks, depth maps or outlines
- JPEG, PNG and WEBP files
About Segment Anything Model 3
Segment Anything Model 3, made by Meta, picks out objects in one image and returns them as masks. A mask is a shape that outlines an object. You can name the object in a text Prompt or mark it with Point Prompts or Box Prompts. A run costs 50 credits.
The image is the only required input. The Prompt box starts with the word "wheel" in it. Apply Mask puts the mask on your image and starts on. Return Multiple Masks starts off. Once it is on, you receive several masks, and Max Masks sets the maximum number, from 1 to 32. It starts at 3. Include Scores returns a number for every mask that shows how sure the model is about it. Include Boxes returns a box around each mask where available. Both start off.
Box Prompts take a box given by its smallest and largest x and y positions, and you can use several boxes. Sam 3 1 costs 100 credits and has a second mode, Image to RLE (run-length encoding, a compact way to store a mask). Segment Anything Model 3 has one mode, Image to Image.
Other models for this job
Same job: find objects and outlines. Compare before you choose.
Image Pro Open
Hi3D Image to ReliefFind objects and outlines
Generate a 3D relief depth map with Hi3D from a single image.
Hitem3D2,000 creditsper run
Image Pro Open
PATINA Image to MapsFind objects and outlines
PATINA creates seamless high-resolution normal, roughness, basecolor (albedo), height (displacement) and metalness maps from images
Open Source100 creditsper run
Image Pro Open
Qwen Image LayeredFind objects and outlines
Qwen-Image-Layered is a model capable of decomposing an image into multiple RGBA layers.
Alibaba500 creditsper run
Image Pro Open
Sam 3 1Find objects and outlines
SAM 3.1 builds comes with Object Multiplex, a shared-memory approach for joint multi-object tracking that delivers faster speeds with larger number of objects tracked.
Meta100 creditsper run
What it does well
Mask preview and mask images
You get a preview of the main mask and the mask images themselves.
Up to 32 masks
With Return Multiple Masks on, Max Masks accepts any number from 1 to 32.
Optional scores and boxes
Include Scores adds a number for each mask that shows how sure the model is about it. Include Boxes adds a box for each mask where available.
How to use Segment Anything Model 3
How It Works
How It Works
01
Upload an image
Add the one image you want to select objects in.
02
Describe or mark the object
Write a Prompt, or give Point Prompts or Box Prompts. The other settings can stay as they are.
03
Run it for 50 credits
Generate and look through the mask preview and mask images.
Tip Only the image is required. Type a Prompt to name the object, because the Prompt box starts with the word "wheel".
Good for
Pick out one object in a photo
Name the object in the Prompt and get a mask that outlines it.
Mark an object by hand
Give points or boxes on the image to show which object you want masked.
Questions
Quick answers to help you get the most out of Segment Anything Model 3 in FunAI Studio.
1. What does Segment Anything Model 3 need from me?
2. What comes back?
3. What is the price?