Garment data for the platforms that need it.
One capture turns a garment into structured data, keyed to your product IDs. LiDAR through the SDK, or standard photos through the API. The measurement API is live today.
{ "external_product_id": "levis-541-32x32", "garment_type": "Trousers", "garment_type_confidence": 0.96, "capture_method": "depth", "processing_ms": 612, "measurements": [ { "name": "Waist Width", "value_cm": 42.5, "confidence": 0.94 }, { "name": "Hip Width", "value_cm": 55.9, "confidence": 0.92 }, { "name": "Front Rise", "value_cm": 29.2, "confidence": 0.91 }, { "name": "Inseam Length", "value_cm": 81.3, "confidence": 0.93 }, { "name": "Leg Opening", "value_cm": 19.7, "confidence": 0.90 } ], "label": { "brand": "Levi's", "model": "541 Athletic Taper", "size": { "primary": "W32 L32", "system": "us" }, "composition": [ { "fiber": "Cotton", "pct": 99 }, { "fiber": "Elastane", "pct": 1 } ], "care": ["Machine wash cold", "Tumble dry low"] }, "description": { "title": "Levi's 541 Athletic Taper Jeans, W32 L32, Dark Wash", "body": "…" }, "media": { "background_removed": "asset_url" }, "version": { "api_version": "v1" }}// Trimmed. Full schema shared on request.One engine. Two ways in.
Every capability runs on the same cloud engine. The only difference is where capture happens: inside your app, or on your server.
- Mobile SDK
- iPhone 12 Pro and newer, iPad Pro 2020+. LiDAR capture, ~5 mm (1⁄5 in) typical.
- Best for listing flows, seller apps, in-store capture, internal warehouse tools.
- REST API
- Any phone or server, standard photos, no LiDAR required. ~9 mm (3⁄8 in) typical.
- Best for warehouses, back-office tools, photo flows.
Everything downstream is identical. One key, one response shape, one billing meter, your product IDs on every record.
What the engine returns.
Point a phone at a garment once. Every capability returns structured JSON your systems can use.
Garment Measurement
Up to 15 points of measure in 0.92 seconds. ~5 mm (1⁄5 in) typical with LiDAR, ~9 mm (3⁄8 in) without.
Label Scanner
Brand, size, fabric mix, care, and origin, read straight off the label into structured fields.
Barcode
The product code from the tag, tied to the same record as the measurements.
AI Descriptions
Ready-to-post listing copy, shaped for each marketplace, in 20 languages.
Ghost Mannequin & Model Shots
Studio imagery from one garment photo. No studio, no crew.
Background removal
Clean cutouts on every capture, edge detail kept.
One record per SKU
Measure, scan, and describe in a single call. Billed per delivered step.
Measurement, background removal, and the combined per-SKU record run as live API endpoints today. Label Scanner, barcode, AI Descriptions, and the imagery modes are delivered through the platform, an integration we build with you and shape to your stack.
Where it plugs in.
Warehouses & resale operations
API · SDK · appCut your intake-to-listing time. One pass measures, scans the label, writes the listing, and shoots the photo. Built for volume.
Marketplaces
SDK · APIMeasurement inside the listing flow. Every item ships with real dimensions.
See what you can build →Brands & retailers
SDK · APIMeasured product data from your studio and merchandising flow. One added step. No new ops.
See what you can build →QA & production teams
The same numbers at every station. A record for every piece, screened against spec.
See garment QC →Custom integrations
APIYour catalog, warehouse, or internal tool, fed by one endpoint. We shape it to your stack.
One key. Everything on it.
- One gateway: api.sizeai.co. Your key carries your entitlements, rate limits, and idempotency.
- Capabilities are settings on your key, not new integrations.
- Named measurements in cm, your product ID on every record, one billable row per result.
- Metered usage. Only delivered results are billed.
- Docs are shared when you request access.
Building a shopping experience on this data? See what brands and marketplaces create with the engine.
For brands and retailers →Built for production scale.
Size AI measures a garment in 0.92 seconds, to ~5 mm (1⁄5 in) typical accuracy with LiDAR.
90+ garment types across 9 categories.
1M+ garments measured for 10,000+ fashion sellers.
Engineering questions we get first.
The shape of objections that come up on every partnership call. If yours isn't here, talk to us.
Integrated into your stack.
Tell us about your platform, your shoppers, and your roadmap. We co-build the right integration with you.
Custom integrations · Reply within one business day
