How to Use SAM 3 image embedding in Windsurf
Returns a base64-encoded image embedding for segmentation workflows. Use it to prepare visual features for later mask generation; the response contains an embedding rather than finished masks.
Prompts to try with SAM 3 image embedding in Windsurf
Copy an example into Windsurf, or ask in your own words.
Use SAM 3 image embedding on https://raw.githubusercontent.com/facebookresearch/sam2/main/notebooks/images/truck.jpg and save the returned response as a fixture for our image preprocessing code.
Use SAM 3 image embedding on https://raw.githubusercontent.com/facebookresearch/sam2/main/notebooks/images/groceries.jpg and add the returned embedding to our annotation prototype's image record.
Use SAM 3 image embedding on https://raw.githubusercontent.com/facebookresearch/sam2/main/notebooks/images/cars.jpg and store the result under a hash of the source image contents.
Use SAM 3 image embedding on https://raw.githubusercontent.com/facebookresearch/sam2/main/notebooks/images/truck.jpg and update our response parser to read the actual returned string.
Use SAM 3 image embedding on https://raw.githubusercontent.com/facebookresearch/sam2/main/notebooks/images/groceries.jpg and add the result to our segmentation experiment manifest, separate from mask outputs.
How to set up SAM 3 image embedding in Windsurf
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Add the server
Open the Customize page from the Agents window and click MCPs. Under This computer, open the config file, or edit it directly: ~/.config/devin/mcp_config.json on macOS and Linux, %APPDATA%\devin\mcp_config.json on Windows. Add this:
{ "mcpServers": { "spiral": { "url": "https://api.spiral.ai/mcp" } } } - Sign in
Click Authenticate on the Spiral server in Customize › MCPs, or let the agent call Spiral once: your browser opens. Sign in to Spiral, pick the organization and click Authorize. No API key.
devin mcp login spiral
- Check it
Spiral is listed under This computer in Customize › MCPs. From the terminal:
devin mcp list
Use the Spiral skill
- Send Windsurf this message
Your agent reads Spiral's skill and sets itself up.
set up https://spiral.ai/SKILL.md
- Sign in or paste a key
Approve Spiral in the browser, or give the agent a key from the console.
Use the Spiral CLI
- Install and sign in
Installs the Spiral CLI, signs you in and adds its skill to your coding agents.
npx -y spiral-cli@latest init
Call the Spiral API
- Create an API key
In the console, under API keys.
- Run SAM 3 image embedding
One request; the result comes back by polling or a webhook.
curl https://api.spiral.ai/v1/runs \ -H "Authorization: Bearer $SPIRAL_API_KEY" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{ "provider": "fal", "model": "fal-ai/sam-3/image/embed", "input": { "image_url": "https://raw.githubusercontent.com/facebookresearch/segment-anything-2/main/notebooks/images/truck.jpg" } }'
Then paste one of the prompts above: Windsurf finds SAM 3 image embedding on Spiral, shows the price and runs it.
SAM 3 image embedding providers and prices
Spiral runs SAM 3 image embedding through fal. Open a card for today's price, with every fee included.
What can you do with SAM 3 image embedding in Windsurf?
Build an annotation tool that prepares SAM 3 image features before users start selecting objects. Keep embedding extraction separate from the code that produces masks.
Prepare images for interactive annotation
Extract visual features before an annotator starts trying segmentation prompts. Keep the embedding available for a compatible downstream mask workflow.
Prepare product photos for cutouts
Generate embeddings for approved product photography before a separate segmentation step selects products and removes backgrounds.
Build a vehicle segmentation test set
Extract embeddings from public vehicle sample images as inputs for later experiments that isolate wheels, body panels, or whole vehicles.
Precompute features for an image archive
Prepare an archive for later segmentation experiments by extracting one embedding per image ahead of the annotation work.
Process new catalog photography
Generate embeddings when approved catalog images arrive or change. Associate each result with the exact image contents used for extraction.
Compare downstream segmentation approaches
Use prepared image features when evaluating compatible mask workflows. Keep feature extraction separate from the experiment that generates and evaluates masks.
More on SAM 3 image embedding and Windsurf
Overview
SAM 3 image embedding in other agents
Questions about SAM 3 image embedding in Windsurf
Can Windsurf use SAM 3 image embedding?
Yes. Add Spiral to Windsurf as an MCP server, and Windsurf can find and run SAM 3 image embedding like any built-in tool.
How much does SAM 3 image embedding cost in Windsurf?
$0.0065 per run at the default inputs. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Windsurf plan is separate.
Do I need a fal account to use SAM 3 image embedding in Windsurf?
No. Spiral runs SAM 3 image embedding on its own fal account. You sign in to Spiral once and never paste a key.
How long does it take to set up SAM 3 image embedding in Windsurf?
About two minutes: add the server, then sign in, then check it.
Does Windsurf wait while SAM 3 image embedding runs?
SAM 3 image embedding runs in the background. Windsurf starts the run, checks on it and gives you the json when it's ready.
What does SAM 3 return for an image?
It returns one base64-encoded image embedding. The published response schema names this string embedding_b64.
Does the result include masks or detected objects?
The embedding response lists no masks, object labels, bounding boxes, or confidence scores. Mask generation is a later segmentation step.
How many images does one call process?
This endpoint processes one image per call. Preparing a collection requires a separate extraction for each image.
Can the same image features support several segmentation attempts?
Yes. fal describes extracting embeddings once and reusing those features for multiple mask prompts without reprocessing the source image.
Does an embedding stay current when the original image changes?
The documented result represents the single image supplied for extraction. It follows that a saved embedding represents those image contents, so a replacement image needs a new extraction.
Start using SAM 3 image embedding in Windsurf
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Windsurf:
set up https://spiral.ai/SKILL.md


