Product Data API
Create a free sandbox key and make your first call in five minutes.
Create a free sandbox key and make your first call in five minutes.
The EcomSource product data API turns any product identifier into a complete product record with one REST call. Send a UPC, EAN, GTIN, ISBN, or ASIN and get back clean JSON: name, brand, images, category, dimensions, every cross-referenced identifier, and Amazon intelligence including variations and buy box data. The sandbox tier is free, 10 calls a day against the full 1.6B+ product index, no card required.
TL;DR
- •One endpoint, every identifier type, JSON out. Built for pipelines, not scraping cleanup.
- •Amazon intelligence included: ASIN conversion, variation mapping, buy box, sales rank.
- •Free sandbox forever, production from $49 per month for 10,000 calls.
- •Bulk CSV and JSON exports plus webhooks on the Professional plan.
What is a product data API?
A product data API is a web service that returns structured information about products on request. Your system sends an identifier, the API answers with the product record as machine-readable JSON. It replaces manual lookups and web scraping with one reliable call, which is why catalog tools, seller software, price trackers, and AI agents are built on top of one.
Your first call
curl -X POST "https://api.ecomsource.ai/api/v1/search/product" -H "X-Access-Key: your_access_key" -H "X-Secret-Key: your_secret_key" -H "Content-Type: application/json" -d '{"identifier": "B07FCZSC41", "identifierType": "asin"}'
Swap identifierType to upc, ean, gtin, or isbn and the same endpoint handles all of them. Full request and response examples in cURL, Node.js, Axios, and Python live in the API documentation.
What comes back
Field group
Contents
Identifiers
UPC, EAN, GTIN, ISBN, and ASIN for the product, cross-referenced
Core record
Title, brand, manufacturer, model, category
Media
Product image links
Logistics
Dimensions and weight where available
Amazon intelligence
Parent and child variations, buy box price, sales rank (plan dependent)
Match metadata
Clean found, ambiguous with candidates, or structured not-found
A response looks like this in outline (illustrative, field names per the live docs): an identifiers block with the upc, ean, gtin, and asin; a product block with title, brand, and category; media with image links; and an amazon block with the variation family and market signals where your plan includes them. One shape, every identifier type, which is what makes client code short.
How to evaluate any product data API, including this one
Run 50 of your own SKUs, not the vendor demo list. Category coverage is the honest differentiator.
Check identifier cross-referencing. A record that returns only what you sent saves you nothing.
Read the miss behavior. Structured not-found beats silent nulls and invented matches.
Price per 1,000 calls at your real volume, free tiers are for testing, not for comparing.
Confirm the commercial license covers your use before building on any provider.
Test the edge formats you actually meet: ISBNs, GTIN-14 case codes, leading-zero UPCs.
What teams build on it
Seller tools: sourcing analysis, listing checks, and repricing inputs keyed by ASIN and UPC.
Catalog and PIM systems: enrich thin records with names, brands, images, and verified GTINs, the pattern in our product database API guide.
Price comparison: match the same product across stores by identifier instead of fuzzy names.
Warehouse software: resolve scans to products at receiving, including case-level codes.
AI agents: ground product questions in identifier lookups, the approach in our piece on product identifier APIs for AI agents.
Batch work without writing a scheduler
Not every job deserves code. A one-off supplier file runs through the bulk barcode lookup with zero programming, and the API picks up where repetition begins. Teams commonly start with the upload tool, prove the match quality on their own catalog, and only then wire the same lookups into their pipeline. The record shape is identical in both paths, so nothing gets rebuilt when you graduate.
API vs scraping, the short version
Scraping product pages gives you HTML that changes weekly, per-site parsers, and legal gray zones. An API gives you one stable contract and data that arrives already structured. If you are weighing the two seriously, our web scraping vs API comparison lays out costs and failure modes without cheerleading.
Pricing, in plain numbers
Figures from the pricing page, current as of August 2026.
Plan
Price
Included
Per 1,000 calls
API Sandbox
$0
10 calls/day, full JSON, full index
Free
API Starter
$49 / month
10,000 calls/month, commercial license
$4.90
API Professional
$149 / month
50,000 calls/month, bulk CSV/JSON exports, webhooks
$2.98
Enterprise
Custom
Millions of calls, database snapshots, SLA
Custom
Limits, errors, and good citizenship
Paid plans meter monthly, so there is no daily cliff mid-batch.
Misses return a structured not-found, never a fake record. Code for nulls on optional fields.
Back off on rate-limit responses rather than hammering, patterns in our API rate limiting guide.
Authentication uses paired access and secret keys, key handling basics in our API authentication guide.
Get your key
The sandbox is free forever: 10 calls a day, full database, no card. Test it against your own products today.
Frequently Asked Questions
Is there a free product data API?
Yes. The EcomSource sandbox is permanently free with 10 calls a day and full JSON responses against the complete 1.6B+ product index. Production plans start at $49 per month for 10,000 calls (August 2026).
Which identifiers does the API accept?
UPC, EAN, GTIN in all lengths, ISBN, and Amazon ASIN, plus name search when you have no code. One endpoint handles all of them via the identifierType field.
Does it support bulk requests?
Yes. Batch through the API on any paid plan, and the Professional plan adds bulk CSV and JSON exports plus webhooks. One-off files can use the bulk barcode lookup instead.
How is this different from scraping product pages?
Scraping returns fragile HTML you must parse per site. The API returns one stable JSON contract, with identifiers cross-referenced, under a commercial license.
Can it convert between ASIN and UPC?
Yes, in either direction, in one call. The response includes every identifier registered to the product plus variation and buy box data on plans with Amazon intelligence.
What data do AI teams use this for?
Grounding. Agents and ML pipelines resolve products by identifier to avoid hallucinated matches, then use the returned attributes as trusted context.
Does it cover products outside the US?
Yes. The index includes international EANs alongside US UPCs, and records cross-reference both. Coverage depth varies by region and category, which is exactly what a 50-product sandbox test on your own catalog reveals.