47 Browser APIs I Use Instead of a Backend

I run 47 PDF tools with zero backend. Here's the browser API arsenal that makes it possible.

File Handling

• File System Access API — open/save files like a native app
• Drag & Drop API — drop PDFs anywhere
• FileReader + Blob — process files without upload

PDF Processing

• pdf.js — render and extract text
• pdf-lib — merge, split, edit, sign
• Tesseract.js — OCR in the browser

Storage & Offline

• Service Workers — full offline PWA
• IndexedDB — store recent files locally
• Cache API — instant repeat loads

AI (the only "server")

• A 200-line Cloudflare Worker proxies Gemini
• Text extracted locally, only text sent to AI
• Canvas renders PDF pages for vision models

Media

• Canvas API — page rendering, image editing
• OffscreenCanvas — heavy work off the main thread

The result: 47 tools, 129 pages, $0/month, works offline.

---

Try it: https://pdfwordtools.pages.dev