Flense

TypeScript/JavaScript client for the Flense document parsing API. Convert PDFs, images, and documents to markdown.

Flense is a paid document parsing service. Get a free demo API key to try it out.

npm install flense

Quick Start

import { Flense } from 'flense';

const flense = new Flense({ apiKey: 'flense_...' });

// Parse a PDF from URL
const result = await flense.parseUrl('https://example.com/doc.pdf').wait();
console.log(result.markdown);

// Parse a local file
import fs from 'fs';
const file = fs.readFileSync('document.pdf');
const result = await flense.parseFile(file, 'document.pdf').wait();
console.log(result.markdown);

Parse Options

All features are OFF by default for fastest processing. Enable them as needed:

const result = await flense
  .parseFile(file, 'document.pdf')
  .withOCR()       // Enable OCR for scanned documents
  .withTables()    // Enable table structure detection
  .withImages()    // Enable image extraction
  .withPageStreaming()  // Stream markdown per-page as each completes
  .wait();

| Method | Description | |--------|-------------| | .withOCR(enabled?) | Enable OCR for scanned/image-based PDFs | | .withTables(enabled?) | Enable table structure detection | | .withImages(enabled?) | Enable image extraction and upload | | .withPageStreaming(enabled?) | Stream markdown per-page as each page completes | | .disableCaching() | Skip cache and force a fresh parse |

Caching

File parse results are cached automatically. When the same file is uploaded with the same options, the cached result is returned instantly with no re-processing. The cache key is a SHA-256 hash of the file content and the parse options.

// Caching is ON by default — identical files return instantly
const result = await flense.parseFile(file, 'doc.pdf').wait();

// Force a fresh parse
const fresh = await flense.parseFile(file, 'doc.pdf').disableCaching().wait();

Caching only applies to parseFile(). URL parsing (parseUrl()) always fetches fresh content since the URL's content may change.

Real-time Progress

For large documents, subscribe to real-time updates:

const job = flense.parseFile(file, 'document.pdf');

job.subscribe({
  onProgress: ({ progress, stage, currentPage, totalPages }) => {
    console.log(`${progress}% - ${stage}`);
  },
  onContent: ({ page, content }) => {
    console.log(`Page ${page} ready`);
  },
  onComplete: (status) => {
    console.log('Done!', status.output?.markdown);
  },
  onError: (error) => {
    console.error('Failed:', error.message);
  },
});

React Hook

import { useParseJob } from 'flense/react';

function DocumentParser() {
  const { parseFile, progress, content, isProcessing, isComplete } = useParseJob({
    apiKey: 'flense_...',
  });

  return (
    <div>
      <input
        type="file"
        onChange={(e) => {
          const file = e.target.files?.[0];
          if (file) parseFile(file);
        }}
      />
      {isProcessing && <p>Processing: {progress?.progress}%</p>}
      {isComplete && <pre>{content}</pre>}
    </div>
  );
}

Environment Variable

Set FLENSE_API_KEY to avoid passing the key explicitly:

export FLENSE_API_KEY=flense_your_key_here

const flense = new Flense(); // API key read from environment

Supported Formats

| Format | Type | |--------|------| | PDF | Documents | | DOCX | Word | | PPTX | PowerPoint | | XLS/XLSX | Excel | | CSV | Spreadsheets | | PNG, JPG, WEBP | Images |

API Reference

See the full API documentation at api.flense.dev/docs

License

MIT