Documentation

Everything you need to get started with PrivDF

Quick Start Guide

1. Open a PDF

Click "Choose PDF" or drag and drop a PDF file anywhere on the page. Your file is loaded into your browser's memory—it never leaves your device.

2. Set a Passphrase

Enter a password in the "Pass-phrase" field. This is used to encrypt your annotations with AES-GCM. Choose something memorable—we can't recover it if you forget.

3. Highlight Text

Select any text in the PDF. A toolbar will appear with "Highlight" and "Add Note" buttons. Click "Highlight" to create a yellow highlight, or "Add Note" to attach a text note.

Keyboard shortcuts: Press H to highlight, N to add a note.

4. Use AI Tools

Click "AI Tools" to open the AI panel. No API key or account is needed — free use includes 5 AI operations per day, and Pro ($4.99/month) raises that limit. Before anything is sent, PrivDF extracts the relevant text locally; your PDF file is never transmitted. You can:

  • Generate document summaries
  • Ask questions about the PDF
  • Get AI-powered insights

5. Export or Delete

Click "Export PDF + Notes" to download your PDF and annotations as an encrypted ZIP file. To delete all your data, click "Delete All Data"—this removes everything from your browser's storage.

What's in the App

Most of the toolset lives behind the Edit PDF button, which opens a dialog with one tab per tool. Everything below runs on your device and is available on the free plan.

Page operations

Merge combines several PDFs into one. Split extracts a page range into a new file. Reorder rearranges pages by dragging. Page Tools inserts blank pages and duplicates existing ones, and pages can be rotated or deleted.

Editing

Edit Region box-selects an area and replaces its contents — it works on scanned pages as well as native text. Edit Text (beta) lets you click directly on words in native-text PDFs. Images places logos and stamps, Watermark adds text watermarks, and PDF forms can be filled in and saved.

Redaction

Redact blacks out a selection permanently. PII Redact scans the document for Social Security numbers, account numbers, addresses, phone numbers and similar identifiers, and lets you review each match before applying. In both cases the underlying content is removed from the file rather than covered with a rectangle.

Signatures

Signature adds a drawn, typed or uploaded signature image. Digital Sign applies a cryptographic PKCS#7/CMS signature so that later modification can be detected. You can import your own .p12 certificate, or generate a self-signed X.509 certificate in the browser — self-signed certificates prove the document hasn't changed since signing, but they do not vouch for your identity the way a certificate from a trusted authority does.

Scanned documents (OCR)

PrivDF detects which pages have little or no extractable text and offers to OCR them on your device. Reading is done locally by Tesseract: free, unlimited, and the page never leaves your computer. The OCR engine is served from PrivDF rather than a third-party CDN, so no outside party sees that you're reading a document; in the desktop app it's inside the bundle, so OCR works with the network switched off from the moment you install. On the web it needs one download of the engine, after which it's cached and works offline too. You get per-page confidence indicators and can re-run any single page. Where on-device reading struggles (dense or multi-column scans), you can re-read that page with AI for higher accuracy; only then is the page image sent, and only that page counts against your AI quota. Pro accounts have unlimited AI, so they default to the more accurate AI read instead — there's a switch in the OCR banner to turn that off and stay on device, and Strict Offline Mode overrides it either way. Exporting a searchable PDF writes an invisible text layer under the original image, so the file stays visually identical but becomes searchable and selectable — including in Adobe Reader.

Getting data out

Drag a box around a table to extract it to CSV, or around an XY chart to pull out its data points. Batch image analysis runs across many pages at once. Annotations export as an encrypted archive you can move to another device.

File size

Compress PDF shrinks the file while keeping the text layer intact. Reduce Size is the aggressive option: it rasterizes pages, which makes the file much smaller but discards selectable text.

Not available yet

Document comparison, Bates numbering and exhibit stamps are not implemented. If one of those is blocking you, say so at support@privdf.ai — it affects what gets built next.

Keyboard Shortcuts

  • H - Create highlight from selected text
  • N - Add note to selected text
  • Esc - Cancel text selection
  • Tab - Navigate between toolbar buttons
  • Enter / Space - Activate focused button

How AI Access Works

You do not need a Venice account or an API key. PrivDF provisions AI capacity for you:

  • Free: 5 AI operations per day, no account required
  • Pro ($4.99/month): a much higher limit, subject to fair use

What actually gets sent: PrivDF extracts text from your PDF on your device, then sends only the passage relevant to your question or summary to Venice.ai, which operates a zero-retention policy. Your PDF file is never transmitted, and nothing is sent until you invoke an AI feature.

Earlier versions of PrivDF asked you to paste your own Venice API key. That is no longer required and the field has been removed.

Troubleshooting

PDF won't load

• Make sure the PDF is not corrupted or password-protected
• Try a different browser (Chrome, Firefox, Edge, Safari all supported)
• Check that JavaScript is enabled
• Clear your browser cache and try again

Highlights disappeared

• Make sure you're using the same passphrase
• Check that you opened the same PDF (same file hash)
• Your browser may have cleared IndexedDB—try importing from a backup ZIP

AI features not working

• Check whether you have used your 5 free AI operations for the day
• Confirm you are online — AI is the one feature that needs a network connection
• Make sure the PDF has extractable text (not scanned images)
• Try a smaller document if you're hitting rate limits

Export/Import issues

• Use the same passphrase for export and import
• Make sure the ZIP file isn't corrupted
• Try exporting again with a fresh export

Browser Compatibility

PrivDF works on all modern browsers that support:

  • Web Crypto API (for encryption)
  • IndexedDB (for local storage)
  • ES2020+ JavaScript (for modern syntax)

Tested browsers:

  • Chrome 120+ ✅
  • Firefox 120+ ✅
  • Safari 17+ ✅
  • Edge 120+ ✅

Need Help?

Can't find what you're looking for?