Skip to main content
Enrich any JSON data with AI-powered web research. No catalog required — just pass your items and define what fields to extract. How it works:
  1. Send items + output schema defining fields to extract
  2. AI searches the web and extracts the specified fields (items processed concurrently)
  3. Get enriched fields with source citations (URLs) for each value

​
Basic request

curl -X POST "https://catalogapi.rastro.ai/api/public/enrich" \
  -H "Authorization: Bearer rastro_pk_..." \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Find product specifications",
    "items": [{"name": "iPhone 15 Pro"}],
    "output_schema": [
      {"name": "price", "type": "string", "description": "Current retail price"},
      {"name": "weight", "type": "string", "description": "Product weight with units"}
    ]
  }'
Response: 
{
  "results": [
    {
      "original_data": {"name": "iPhone 15 Pro"},
      "enriched_fields": {
        "price": {
          "value": "$999",
          "sources": ["https://apple.com/shop/buy-iphone/iphone-15-pro"]
        },
        "weight": {
          "value": "187 grams",
          "sources": ["https://apple.com/iphone-15-pro/specs"]
        }
      },
      "all_sources": ["https://apple.com/..."],
      "error": null
    }
  ],
  "total_items": 1,
  "successful": 1
}

​
Parameters

ParameterRequiredDefaultDescription
promptYes-Search prompt for web enrichment
itemsYes-List of JSON objects to enrich
output_schemaYes-List of field definitions: [{name, type, description}, ...]
allowed_domainsNoanyRestrict web sources to these domains only
speedNo"medium""fast", "medium", or "slow"
Field types: string, number, integer, boolean, array

​
Multiple items

Items are processed concurrently: 
curl -X POST "https://catalogapi.rastro.ai/api/public/enrich" \
  -H "Authorization: Bearer rastro_pk_..." \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Find product specifications",
    "items": [
      {"name": "iPhone 15 Pro"},
      {"name": "Samsung Galaxy S24"}
    ],
    "output_schema": [
      {"name": "price", "type": "string", "description": "Current retail price"},
      {"name": "display", "type": "string", "description": "Screen size and type"},
      {"name": "battery", "type": "string", "description": "Battery capacity"}
    ]
  }'

​
Domain restrictions

Restrict web sources to specific domains: 
curl -X POST "https://catalogapi.rastro.ai/api/public/enrich" \
  -H "Authorization: Bearer rastro_pk_..." \
  -H "Content-Type: application/json" \
  -d '{
    "prompt": "Find official specs",
    "items": [{"name": "MacBook Pro 16"}],
    "output_schema": [
      {"name": "cpu", "type": "string", "description": "Processor model"},
      {"name": "ram", "type": "string", "description": "Memory options"},
      {"name": "price", "type": "string", "description": "Starting price"}
    ],
    "allowed_domains": ["apple.com"]
  }'

​
Python example

import requests

API_KEY = "rastro_pk_..."
BASE_URL = "https://catalogapi.rastro.ai/api"

response = requests.post(
    f"{BASE_URL}/public/enrich",
    headers={
        "Authorization": f"Bearer {API_KEY}",
        "Content-Type": "application/json"
    },
    json={
        "prompt": "Find product specifications",
        "items": [
            {"name": "Sony WH-1000XM5"},
            {"name": "Apple AirPods Pro 2"}
        ],
        "output_schema": [
            {"name": "price", "type": "string", "description": "Current retail price"},
            {"name": "weight", "type": "string", "description": "Product weight"},
            {"name": "battery_life", "type": "string", "description": "Battery duration"}
        ]
    }
)

result = response.json()
print(f"Processed {result['total_items']} items, {result['successful']} successful")

for item_result in result["results"]:
    print(f"\n{item_result['original_data']['name']}:")
    for field, data in item_result["enriched_fields"].items():
        print(f"  {field}: {data['value']}")
        print(f"    Sources: {data['sources']}")
Need more control? Use Flows to build reusable pipelines with transformations, filters, and multi-step enrichment.
-> Full API reference