Catalog raises $3M to build the product data layer for AI commerce. Read the announcement.

The product data layer
for agentic commerce.

Live, structured data for any product on the open web.

Catalog API playground: configure a product Extract request and view the generated Python, Javascript, or curl code side-by-side.

Any product on the internet.

Drop a URL or a domain — Shopify, Amazon, custom builds, raw HTML. Get back the same 86+ normalized fields. No scrapers to maintain.

The primitives every commerce agent needs.

Six kinds of builders. One set of primitives. Built for the systems that transact — not the demos.

  1. Shopping agents

    Ground every answer in real products, prices, and stock — not a model's best guess.

    • grounding
    • live pricing
    • stock confidence
  2. Product search

    Index the open commerce web instead of a single feed. Stable IDs make embeddings and rerankers worth maintaining.

    • open-web index
    • stable IDs
    • reranking
  3. Recommendations

    Pull live inventory and price before serving. Stop recommending what isn't in stock.

    • inventory check
    • price-aware
    • personalization
  4. Comparison

    Variant-level parity across merchants. Match SKUs with consistent attributes and units.

    • variant match
    • cross-retailer
    • normalized units
  5. Marketplaces & affiliate apps

    Real-time assortments and monetizable outbound links. Affiliate commissions where supported, opt-in per request.

    • live assortments
    • outbound links
    • opt-in affiliate
  6. Research labs

    Reproducible product objects with timestamped freshness. Useful ground truth for benchmarking agents.

    • timestamped
    • reproducible
    • ground truth

Why teams switch to Catalog.

Five common paths to product data — and the trade each one quietly asks you to keep paying. We built Catalog so you stop paying it.

Alternative
The reality
In-house scrapersBuild & maintain

Break on every redesign. Return HTML.

We absorb the bot defenses, retries, and schema drift. You get typed objects.

Merchant feeds & PIMsStatic catalogs

Partial coverage. Missing the depth AI needs.

Any merchant on the open web, live, with the depth agents require.

Shopping SERP APIsSearch wrappers

Links and snippets. No variants or specs.

Full product objects. Variant-level price and stock. Normalized specs.

Affiliate networksLink inventory

Built for links, not data.

Product-first. Affiliate is an option, not the model.

Building it yourselfCustom stack

Months to build. Maintenance forever.

One integration. Your team works on the product.

Modular pricing per call.

Modular, per-listing pricing. No minimums. Volume discounts kick in automatically.

Pay as you go

Per-call pricing. No commitments.

Crawl

Capture all URLs of a store.
  • ID
  • URL
  • Name
  • Images
  • Price
  • Vendor
$0.2
/100 listings
Extract

Flexible extraction depth per product URL.

Starter

  • Pricing and variant availability
  • Image URLs
  • GTIN, brand, category
  • Aggregated review count and rating
  • Full HTML
$2
per 100

Scale

Includes Starter
  • AI enrichment
  • Reddit insights
  • Aggregated reviews
$10
per 100

Pro

Includes Scale
  • Image analysis
  • Review analysis
  • Similar products
  • Dynamic FAQs
  • Deep research
$100
per 100
Volume or private deployment?Talk to us

Questions builders ask.

Any e-commerce site on the open web. No partner agreements, no merchant onboarding, no gated coverage.

Ship the agent. We’ll handle the data.

One API. Any merchant. Live in an hour.