▤
PDF Annotator
No PDF open
/ 0

Open or drop a PDF

Read technical PDFs, annotate evidence, and export review notes as an annotated PDF or Markdown package.

Privacy: PDFs stay in your browser unless you run OCR or table extraction. See Help for details.
Help and app guide Read, annotate, extract selected content, and export PDFs.

What this app is for

Use the app to study or review PDFs, mark important content, extract selected regions, and export the result as an annotated PDF or Markdown package.

1OpenChoose, drop, or reopen a PDF.
2ReadNavigate pages, bookmarks, links, and text.
3Annotate or extractMark text and figures or extract a selected region.
4ExportSave an annotated PDF or Markdown package.

Reading and navigation

Move through the PDF

Use the page box, previous/next buttons, custom zoom, or Fit width. Browser Back and Forward return to earlier PDF positions.

Bookmarks and links

Contents shows embedded bookmarks. Internal links jump within the PDF; external links open in a new tab while Select is active.

Select and copy text

Use Select (V) to copy text or move annotations. Use Highlight (H) to turn selected text into a PDF highlight.

Recent files

Open keeps up to 10 source PDFs in this browser, up to 512 MB total. Clear them from the same menu.

Annotations and notes

Highlights and image regions

Create highlights, rectangles, ellipses, and concave polygons. Hold Ctrl for a square or circle; press Enter or double-click to finish a polygon.

Edit existing annotations

With Select active, move regions or reshape polygon vertices. The inspector edits the title, note, color, opacity, and Markdown inclusion.

Sticky notes

Note (N) creates a PDF text annotation. Add its title and comment in the inspector.

Markdown note preview

Raw and Rendered views support GitHub-flavored Markdown and KaTeX using $…$, $$…$$, \(…\), and \[…\]. Code remains literal.

Captures and AI-assisted extraction

High-resolution screenshot

Screenshot (S) exports the selected region as a high-resolution PNG independent of the current viewer zoom.

OCR to Markdown

OCR (O) sends the selected crop to the configured provider and returns Markdown with LaTeX for review.

Table extraction

Table (T) sends the selected crop and returns a Markdown table. Recommended reasoning uses low effort for OCR and medium for tables where supported.

Adding an extraction

Add to document creates a rectangle annotation containing the reviewed Markdown. Markdown export also includes the source crop.

Export, saving, and compatibility

Annotated PDF

Exports supported annotations as native PDF objects rather than flattening them into the page image.

Markdown ZIP or folder

Creates one Markdown file plus an assets/ folder of PNG crops. Folder export is available in supported browsers.

Imported PDF features

Imports editable Highlight, Square, Circle, Polygon, and Text annotations. Bookmarks and links remain available for navigation.

What is remembered

Recent files store the source PDF only. Settings and sidebar widths persist, but annotation changes must be exported.

Important: Recent files do not save annotation changes. Export before closing, reloading, or opening another PDF.

Keyboard shortcuts

Open, undo, and redo
Ctrl/Cmd+O open PDF · Ctrl/Cmd+Z undo · Ctrl/Cmd+Y or Ctrl/Cmd+Shift+Z redo
Zoom
Ctrl/Cmd++ zoom in · Ctrl/Cmd+- zoom out · Ctrl/Cmd+0 fit width · Ctrl/Cmd+wheel/pinch zoom around pointer
Choose a tool
V select · H highlight · R rectangle · E ellipse · P polygon · N note · S screenshot · O OCR · T table
Complete or cancel drawing
Enter finish polygon · Escape cancel drawing, close a mobile panel, or clear the current selection state
Delete selected annotation
Delete or Backspace
Sidebars
Alt+[ toggle left sidebar · Alt+] toggle right sidebar
Resize a focused sidebar divider
Arrow key adjust by 8 px · Shift+Arrow adjust by 24 px · Home/End minimum/maximum · double-click to reset

Privacy and data flow

Local browser processing

PDF viewing, annotation, screenshots, Markdown and KaTeX preview, and exports run in the browser.

Code and font downloads

The app loads its libraries and worker from cdnjs, and IBM Plex fonts from Google Fonts. PDF content is not attached to those requests.

AI extraction requests

Only OCR or Table extraction sends data: the selected crop, request, model settings, credentials, and extra headers go to the configured endpoint. The full PDF is not sent.

Stored on this device

Recent PDFs use IndexedDB. Settings and provider configuration use local storage. API keys are stored only when Remember the key is enabled.

The CDN and selected AI provider follow their own policies. Clear Recent files and browser site data to remove local copies and settings.

PDF Annotator & Markdown Exporter Version 2.16
Settings

Image rendering

px

Default: 1024 px.

px

Higher values improve small text but increase request size. Default: 1600 px.

AI extraction provider

Recommended uses low effort for OCR and medium effort for table reconstruction.

API keys are sent only to the configured endpoint when an extraction runs. The app has no server component or subscription OAuth flow. Keys stay in memory unless “Remember the key” is enabled; do not enable it on a shared device.
Extraction result

Review the model output. Adding it to the document creates a native rectangle annotation whose comment contains this Markdown; Markdown export also includes the source crop as a separate PNG.
Working…