Page Extract API

Extract the data out of any PDF.

Describe the fields you want in a JSON Schema; get them back from the document. Long filings, scans and tables of thousands of rows included.

Get API key

Powered by PageIndex Agent

renaissance_13f.pdf

{
"holdings": [
{
"issuer": "111 INC",
"value": 84851
},
… 3,062 more
]
}

The most accurate on ExtractBench, for less

Page Extract against other extraction methods on ExtractBench: accuracy, and what each costs a page.

96.80overall value F1 with Pro, the highest on ExtractBench
$0.03a page with Standard, which still scores above every other method
60%less a page than LlamaExtract Agentic Plus, the most accurate of the others

Accuracy and price

Page Extract Other methods

Page Extract pro96.80 · $0.05 / page
Page Extract standard96.37 · $0.03 / page
LlamaExtract Agentic Plus96.17 · $0.075 / page
Reducto Deep Extract96.08 · $0.0635 / page
Pulse (Effort)96.07 · $0.105 / page
LlamaExtract Agentic95.65 · $0.0312 / page
Codex (GPT-6.1 Sol)95.31 · $0.0713 / page

Overall value F1 (axis starts at 94)

Show all scores
MethodOverallShortMediumLongPer page
Page Extract pro96.8097.4695.4894.86$0.05
Page Extract standard96.3796.9894.9995.46$0.03
LlamaExtract Agentic Plus96.1797.0294.3094.65$0.075
Reducto Deep Extract96.0896.8494.8092.85$0.0635
Pulse (Effort)96.0796.8394.7093.24$0.105
LlamaExtract Agentic95.6596.0694.7994.73$0.0312
Codex (GPT-6.1 Sol)95.3197.0192.9685.46$0.0713

How it works

1 Your PDF
13f_holdings.pdf
Name of issuerValue ($)
10X GENOMICS INC21,415,529
111 INC84,851
17 ED & TECHNOLOGY GROUP INC83,779
3,060 more rows
2 The fields you want
schema.json
{"type": "object",
 "properties": {"holdings": {
   "type": "array",
   "items": {"type": "object",
     "properties": {
       "issuer": {"type": "string"},
       "value_usd": {"type": "number"}
}}}}}
3 Every row, as JSON
result.json
{"holdings": [
  {"issuer": "10X GENOMICS INC",
   "value_usd": 21415529},
  {"issuer": "111 INC",
   "value_usd": 84851},
  // … 3,061 more
]}

How to write a schema →

Pay for the pages you extract

Paid when a job completes; a cancelled or failed job is free. The same credit as Page Search, and $10 of it free to start.

Standard

$0.03 / page

For most documents.

Pro

$0.05 / page

A stronger model, for the hardest layouts.

Try it on your own documents.

Get API key