Structured data for any URL on the web

Schema.org for any URL — even sites that don't have it

Most of the web has no structured data. Our API generates it anyway. Point it at any public page — yours, a competitor's, a supplier's, a source you're aggregating — and get clean, validated JSON-LD back in under two seconds.

What developers build with it

Rich search experiences. Comparison platforms. AI agents. Aggregators. Internal tooling. Anywhere structured data is the blocker between raw web content and the product you want to ship.

Three endpoints, one API key

The API is intentionally minimal. Most integrations use one endpoint. Authentication is a single header. Responses are plain JSON-LD — no custom envelopes, no SDK required.

See it work

Four request/response pairs showing common use cases. Copy and paste any of them — they work on your free tier right now.

# Request
curl -X POST https://api.markupschema.com/api/ms/schema/generate \
  -H "X-API-Key: bl_your_key" \
  -d '{"url": "https://example-restaurant.ie"}'

# Response
{
  "@type": "Restaurant",
  "name": "The Wild Kitchen",
  "address": { "streetAddress": "42 Grafton St", "addressLocality": "Dublin" },
  "openingHours": ["Mo-Fr 12:00-22:00", "Sa-Su 10:00-23:00"],
  "priceRange": "$$",
  "aggregateRating": { "ratingValue": 4.6, "reviewCount": 284 }
}

# Request
curl -X POST https://api.markupschema.com/api/ms/schema/generate \
  -H "X-API-Key: bl_your_key" \
  -d '{"url": "https://store.example.com/products/wool-throw"}'

# Response
{
  "@type": "Product",
  "name": "Donegal Wool Throw",
  "brand": { "name": "McNutt of Donegal" },
  "offers": {
    "price": 89.00,
    "priceCurrency": "EUR",
    "availability": "https://schema.org/InStock"
  },
  "aggregateRating": { "ratingValue": 4.8, "reviewCount": 62 }
}

# Request
curl -X POST https://api.markupschema.com/api/ms/schema/generate \
  -H "X-API-Key: bl_your_key" \
  -d '{"url": "https://news-site.com/schema-in-2026"}'

# Response
{
  "@type": "Article",
  "headline": "The State of Schema.org in 2026",
  "author": { "name": "Sarah Chen" },
  "datePublished": "2026-04-15",
  "dateModified": "2026-04-18",
  "publisher": { "name": "Tech Weekly" },
  "image": "https://news-site.com/img/schema-2026.jpg"
}

# Request
curl -X POST https://api.markupschema.com/api/ms/schema/batch \
  -H "X-API-Key: bl_your_key" \
  -d '{"urls": ["https://site.com/p/1", "https://site.com/p/2", "...up to 100"]}'

# Response
{
  "job_id": "batch_abc123",
  "status": "queued",
  "url_count": 100
}

# Poll: GET /api/ms/schema/status/batch_abc123

Plug schema into your AI tools, natively

Model Context Protocol is how modern AI tools — Claude, ChatGPT Desktop, Cursor — discover and use external capabilities. Our MCP server exposes schema generation as a tool call.

{
  "mcpServers": {
    "markupschema": {
      "url": "https://api.markupschema.com/mcp",
      "headers": { "X-API-Key": "bl_your_key" }
    }
  }
}

Add this to Claude Desktop's config, Cursor, or any MCP client. Your AI now has on-demand access to schema.org data for any URL on the web. No glue code. No parsing. No hallucinated data.

Numbers that matter in production

Responses are cached per-URL for 24 hours by default — override with fresh: true for live regeneration. Failed generations retry automatically up to three times. Every response includes a cached boolean and an inference_ms field.

Straightforward pricing, no metered surprises

Every paid tier has a daily request limit, not a monthly one — burst traffic doesn't blow your budget. Cached responses don't count against your limit.

Authentication and errors

All API requests require an API key passed as an X-API-Key header. Keys are created and rotated in your developer console.

All error responses include a JSON body with an error code, a human-readable message, and (where applicable) a retry_after seconds hint.

Use it from any language

The API is plain JSON over HTTPS — no SDK required. Any HTTP client works. No custom auth flow, no WebSocket handshake, no proprietary encoding.

Everything you need to go live

Common questions from developers