The Problem with PDF Editing
Adding visual elements to PDFs traditionally requires jumping between multiple tools. You need design software to create graphics, image editors to prepare assets, and then PDF tools to assemble everything. It's time-consuming and requires skills many people don't have.
FoundryPDF takes a radically different approach: just describe what you want.
Natural Language PDF Editing
Instead of learning complex tools, you work in plain English. Want to add a historical portrait to your presentation? A diagram explaining a concept? A decorative element that matches your document's style?
Just say so.
The AI understands context from your existing document, uses style references from pages you select, and can even search for current information when needed. The result is seamlessly integrated into your PDF with the text layer preserved.
How It Works
1. Upload your PDF - Drag and drop or use the file picker. The visual viewer shows thumbnails of all pages with zoom controls.
2. Select pages to edit - Click on the pages you want to modify. You can select multiple pages for batch editing.
3. Configure options - Choose your resolution tier, enable PDF context analysis for smarter edits, and optionally select style reference pages.
4. Describe your changes - Write what you want in plain English. Be as specific or creative as you like.
5. Watch it happen - Real-time progress updates show exactly what's being processed via WebSocket connections.
6. Download the result - Get your enhanced PDF with all modifications applied and text layer intact.
Technical Architecture
FoundryPDF runs on a Python backend powered by FastAPI, chosen for its async capabilities and type safety with Pydantic. PDF processing uses a combination of specialized libraries: pdf2image for rendering, pypdf for manipulation, and Tesseract for OCR when needed.
The AI magic comes from Google's Gemini model, specifically leveraging its image generation capabilities to create contextually appropriate visuals based on your prompts.
The React frontend provides a responsive, modern interface with real-time updates. Authentication is handled through Better Auth, and payments run through Stripe with a credit-based model that lets users pay for exactly what they need.
Credit System
Pricing is transparent and usage-based:
| Quality Tier | Resolution | Credits/Page | |--------------|------------|--------------| | Standard | 1K | 1 credit | | High Definition | 2K | 2 credits | | Ultra HD | 4K | 3 credits |
Credit packs start at $10 for 10 credits, with volume discounts available for larger purchases.
Who It's For
FoundryPDF is built for anyone who needs to enhance documents without design expertise:
- Authors adding illustrations to ebooks and manuscripts
- Presenters enhancing slides with custom graphics
- Educators creating visually engaging course materials
- Professionals polishing reports and proposals
The goal isn't to replace design tools - it's to eliminate the barrier between having an idea and seeing it in your document.