Getting started

Use whole File, Blob, ArrayBuffer or Uint8Array input; Node Buffer works too. Path strings and remote range requests are outside the input contract. Node applications can read a file first, then pass its bytes to "@pdfextract/core".openPdf.

pnpm add @pdfextract/core @pdfextract/ocr @pdfextract/storage
Copy

The libraries support browser ESM and Node ≥22.12.0. Configure browser workers using self-hosted assets before opening a PDF in a bundled application.

Native text and images

This typechecked Node example returns owned exports that remain usable after close:

import { openPdf, type PdfInput } from '@pdfextract/core';

/** Extract native text and independent image exports from a whole PDF. */
export async function extractNative(input: PdfInput) {
  const pdf = await openPdf(input);
  try {
    const text = await pdf.getStructuredText({ ocr: 'off' });
    const exports = [];
    for (const image of await pdf.getImages()) {
      const full = await pdf.extractImage(image.id);
      const thumbnail = await pdf.extractImage(image.id, { maxWidth: 320, maxHeight: 240 });
      exports.push({ image, full, thumbnail });
    }
    return { text, exports };
  } finally {
    await pdf.close();
  }
}
Copy

For large documents, upload each full image and thumbnail inside the loop instead of accumulating all exports in an array. The engine decodes on demand; the application controls the lifetime of the resulting bytes.

The document's four operations are getStructuredText, getImages, extractImage and close. The structured result includes pages, blocks, lines, spans, provenance, warnings and derived fullText. It round-trips through JSON.

Run the complete examples

From the repository root, run pnpm build && pnpm assets && pnpm dev and open http://127.0.0.1:4173/pdfextract-demo/. The default browser adapter uses IndexedDB. Reload the page and load the saved extraction to retrieve its manifest and thumbnails without parsing the PDF again. Select an image to receive the stored full-size File.

PDFEXTRACT_LANGUAGE_PATH="$PWD/node_modules/@tesseract.js-data/eng/4.0.0" \
  pnpm example:node extract tests/fixtures/mixed.pdf ./pdf-assets
pnpm example:node load '<printed-manifest-key>' ./pdf-assets
Copy

The source examples are in the repository's browser and Node directories.