Skip to main content
POST
Extract text

Phase

Phase 2 — Text extraction (non-OCR). Uses embedded text layers.

Request

object
required
The PDF to extract from. One of:
  • { "pdf_id": "pdf_abc123" }
  • { "url": "https://example.com/doc.pdf" }
  • { "base64": "JVBERi0x..." }
string
default:"all"
Which pages to extract. "all", "1-3", "1,3,5".
string
default:"text"
Output format:
  • text — Plain text, pages separated by \n\n---\n\n.
  • markdown — Best-effort markdown conversion (headings, lists, tables).
  • structured — JSON with per-page text, bounding boxes, and reading order.

Response

text format

structured format

This endpoint extracts text from the PDF’s text layer. For scanned documents or images, use the OCR endpoint.

Example