How it works
Upload an image containing one or more items, and Vision:
- Extracts item details from the photo (name, description, materials, categories)
- Optionally processes the extracted details to:
- Assign HS codes via Classify
- Determine manufacturing origin via Country of Origin
- Estimate monetary value via Customs Value (doc coming soon)
Vision can detect multiple items in a single photo—each item is returned as a separate object in the response. All of this happens in a single API request with response times averaging 2-5 seconds, including image upload and all AI analysis.
Use cases
Customs agencies: Expedite physical inspections by photographing package contents to instantly extract item details for verification against declarations, reducing manual documentation time and enabling faster clearance decisions while maintaining inspection accuracy.
Package forwarders: Photograph an incoming package to capture item names, descriptions, materials, and categories instantly—no need to manually type details for every item. Enable optional classification and valuation for customs compliance without requiring specialized trade knowledge from intake staff.
Customs brokers: Document shipment contents with AI-extracted details and optional HS code assignment, reducing manual assessment time from minutes to seconds while maintaining declaration accuracy.
Logistics operations: Process high volumes consistently by automating item data capture. Enable classification for customs-critical shipments while keeping simple domestic packages fast with extraction-only mode.
Getting started
API guide
Vision is available exclusively via API. Use the itemsExtract mutation to upload an image and request item detection along with optional classification, country of origin, and customs value estimates.
mutation ItemsExtract($input: ItemsExtractInput!) { itemsExtract(input: $input) { id quantity content { categories description language materials name } classification { id confidenceScore hsCode { code } } countryOfOriginInference { id confidenceScore countryOfOrigin } valueEstimation { id currency value valueEstimateRange { high low width } } }}Parameters:
configuration: Toggle classification, value estimation, and country of origin inference- Options:
"ENABLED"or"DISABLED"
- Options:
imageBase64: Base64-encoded image data (JPEG, PNG, or HEIC)localizedLanguageCode: Language code for item descriptions (e.g.,"EN","ES")shipToCountry: Destination country code for context
Response fields:
Each detected item includes:
content: Name, description, materials, and categories in the requested languageclassification: HS code with confidence score (when enabled)countryOfOriginInference: Manufacturing origin with confidence score (when enabled)valueEstimation: Estimated value and plausible range (when enabled)
Access and pricing
Zonos Vision requires a subscription. To access Vision, contact sales.
Zonos Vision
Transform photos into customs-ready shipment data.
Zonos Vision extracts item details from photos using AI—capturing names, descriptions, materials, and product categories automatically. This eliminates manual data entry for item documentation.
You can then optionally process these extracted details to classify items with HS codes, determine country of origin, and estimate customs monetary values. All of this happens in a single API call.
Traditional shipment processing requires trained staff to manually describe, classify, value, and source items—a time-consuming process demanding expertise in customs regulations and product knowledge. Vision transforms this: take a photo, review the extracted details, and ship. This reduces processing time, lowers labor costs, and makes international shipping accessible to less specialized staff.