Playground

Drop a PDF in. Get structured data out.

The live /v1/extract endpoint. No OCR and no model — run the same file twice and you get the same answer. Looking for HTML to PDF?

Shared and public — 5 calls an hour per IP, no quota used.

Input

<style>
  body { font: 12px Helvetica, sans-serif; margin: 40px; }
  .label { color:#666; display:inline-block; width:110px; }
  table { width:100%; border-collapse:collapse; margin-top:24px; }
  th { text-align:left; border-bottom:2px solid #111; padding:7px 6px; font-size:11px;
       text-transform:uppercase; }
  td { border-bottom:1px solid #ddd; padding:7px 6px; }
  .num { text-align:right; }
</style>
<h1>INVOICE</h1>
<div><span class="label">Invoice Number:</span> INV-2026-0417</div>
<div><span class="label">Invoice Date:</span> 21/08/2026</div>
<div><span class="label">Status:</span> pending</div>
<table>
  <tr><th>Description</th><th class="num">Qty</th><th class="num">Unit Price</th><th class="num">Total</th></tr>
  <tr><td>Professional Services</td><td class="num">10</td><td class="num">1,000.00</td><td class="num">10,000.00</td></tr>
  <tr><td>Laptop — MacBook Air M4</td><td class="num">2</td><td class="num">50,000.00</td><td class="num">100,000.00</td></tr>
  <tr><td>Onboarding &amp; setup</td><td class="num">1</td><td class="num">5,000.00</td><td class="num">5,000.00</td></tr>
</table>

Schema — optional

Leave it empty to get every Label: value on the page.

Output

The extracted tables and fields appear here.