# Nano PDF
Nano PDF is a CLI tool that edits PDF slides using natural language prompts, powered by Google's Gemini 3 Pro Image model. It can modify existing slides or generate new ones that match the visual style of the deck.
## Key Features
- **Natural language editing**: change titles, update graphs, fix typos across multiple slides
- **Add new slides**: generate slides matching the existing deck's visual style
- **Non-destructive**: preserves searchable text layer via OCR re-hydration (Tesseract)
- **Multi-page and parallel**: edit multiple pages in a single command with concurrent processing
- **Configurable resolution**: 4K (default), 2K, 1K to balance quality vs cost
- **Google Search integration**: model can search the web for information before generating
- **Style references**: specify pages to use as style reference for consistent visual output
## How It Works
1. **Page rendering**: converts target PDF pages to images using Poppler
2. **Style references**: optionally includes reference pages for fonts, colors, layout
3. **AI generation**: sends images + prompts to Gemini 3 Pro Image
4. **OCR re-hydration**: Tesseract restores searchable text layer
5. **PDF stitching**: replaces original pages with AI-edited versions
## Usage
```bash
pip install nano-pdf
nano-pdf edit <file> <page> "<prompt>"
nano-pdf add <file> <position> "<prompt>"
```
Requires Poppler, Tesseract, and a paid Google Gemini API key (free tier does not work).
## Technical Details
- Written in Python (3.10+)
- MIT license
- 1.2k GitHub stars
## References
- https://github.com/gavrielc/Nano-PDF
## Related
-