🪄 HyperPaper

Next-Generation Intelligent PDF Reader and Annotation Tool

Making paper reading simple, efficient, and intelligent

Features • Quick Start • Demo • Tech Stack • Configuration • Contributing

Language / 语言: 🇬🇧 English | 🇨🇳 中文

✨ Why Choose HyperPaper?

When reading academic papers, have you ever encountered these problems?

📚 Complex Formulas - Complex mathematical formulas require constant reference to materials
🌐 Language Barriers - Foreign papers are difficult to understand, requiring frequent switching of translation tools
📝 Chaotic Annotations - Switching between multiple tools makes annotations difficult to manage uniformly
🔍 Understanding Difficulties - The meaning of charts and tables requires additional queries

HyperPaper solves all these problems for you!

HyperPaper is an intelligent PDF reader designed specifically for academic research, integrating AI Q&A, OCR recognition, formula parsing, intelligent translation, and a powerful annotation system, making paper reading unprecedentedly efficient.

🚀 Core Features

✨ AI-Powered Intelligent Q&A

Regional Q&A: Select any area, directly translate or ask questions, AI answers for you
Multi-Model Support: Supports Qwen series and other AI models
Context Understanding: Provides accurate answers based on selected content
Markdown Rendering: Supports LaTeX formulas, code blocks, and rich formats

📸 Powerful OCR Capabilities

Local OCR: Local recognition engine based on Pix2Text, protecting privacy
Formula Recognition: Automatically recognizes mathematical formulas and converts them to LaTeX
Chart Extraction: Intelligently extracts text and structure from charts
Real-time Progress: OCR processing progress is visualized

🌍 Intelligent Translation

Multi-language Support: Chinese, English, Japanese, Korean, French, German, Spanish
Auto Detection: Intelligently recognizes source language
Target Language Selection: Customize translation target language in preferences
Silent Updates: Translation results are automatically updated without manual refresh

✏️ Rich Annotation System

Text Annotation: Highlight, underline, strikethrough
Free Drawing: Hand-drawn annotations, as you wish
Note Function: Click to add notes, supports multi-line editing
Text Comments: Add text descriptions directly on PDF
Color Customization: Rich color selection, personalized annotations
Secondary Editing: All annotations support editing and deletion

🎨 Modern UI Design

Liquid Glass Style: Semi-transparent liquid glass effect, visually elegant
Floating Toolbar: Doesn't block content, convenient operation
Smooth Animation: Silky smooth interaction experience
Responsive Layout: Adapts to different screen sizes

📊 Formula and Chart Processing

Formula Recognition: Three processing modes (no formula processing, local OCR + LLM API translation, VLM API translation)
LaTeX Rendering: Perfect support for mathematical formula display
Chart Understanding: AI analyzes chart content and provides explanations

🎬 Demo

🤖 AI Features Demo

Watch how HyperPaper's AI-powered features work:

https://github.com/Mengqi-Lei/HyperPaper/releases/download/Demo-video-1080p/hyperpaper-AI.mp4

Features shown: Regional Q&A, OCR recognition, intelligent translation, and formula processing

✏️ Annotation Features Demo

See HyperPaper's powerful annotation system in action:

https://github.com/Mengqi-Lei/HyperPaper/releases/download/Demo-video-1080p/hyperpaper-notes.mp4

Features shown: Text annotation, highlighting, note-taking, and annotation management

Quick Feature Overview

Regional Q&A

1. Select any area in the paper
2. Enter questions in the right-side Q&A panel
3. AI provides accurate answers based on selected content

OCR Recognition

1. Select areas containing formulas or charts
2. Automatically triggers OCR recognition
3. Recognition results are automatically displayed, supporting translation and explanation

Intelligent Annotation

1. Select annotation tools (highlight/underline/strikethrough/draw/note/text)
2. Annotate on PDF
3. All annotations are automatically saved, supporting secondary editing

🛠️ Tech Stack

Frontend

SwiftUI - Modern UI framework
PDFKit - PDF rendering and interaction
AppKit - macOS native components

AI Services

Qwen API - Large language model service
Pix2Text - Local OCR engine
Vision API - Formula and chart recognition

Core Features

PDF Annotation System - Complete PDF annotation support
Markdown Rendering - Supports LaTeX formulas
Multi-language Translation - Intelligent translation engine

📦 Quick Start

System Requirements

macOS 12.0 or higher
Xcode 14.0 or higher (development environment)

Installation Steps

Clone Repository

git clone https://github.com/Mengqi-Lei/HyperPaper.git
cd HyperPaper

Open Project
```
open HyperPaper/HyperPaper.xcodeproj
```
Configure API Key ⚠️ Required Step
- Open HyperPaper/HyperPaper/Models/APIConfig.swift
- Replace YOUR_API_KEY_HERE with your actual API Key
- For detailed configuration instructions, see API Configuration Guide
- Get API Key: https://api.probex.top
Build and Run
- Select target device in Xcode
- Press Cmd + R to run the project

First Use

Open PDF
- Click "Select PDF File" button
- Or use menu bar File > Open
Start Reading
- Use toolbar to switch between different modes (reading/annotation)
- Select areas for Q&A or OCR
- Use annotation tools for marking
Configure Preferences
- Open preferences (Cmd + ,)
- Select AI model
- Set formula processing mode
- Choose translation target language

🎯 Use Cases

📖 Academic Paper Reading

Quickly understand complex formulas
Translate foreign papers
Record reading notes
Organize key information

📚 Literature Review

Batch process multiple papers
Unified annotation management
Extract key content
Generate reading summaries

🔬 Research and Learning

Deeply understand chart meanings
Analyze experimental data
Compare different viewpoints
Build knowledge systems

⚙️ Configuration

API Key Configuration (Required)

Before using AI features, you need to configure API Key:

Open HyperPaper/HyperPaper/Models/APIConfig.swift
Replace YOUR_API_KEY_HERE with your actual API Key
Get API Key: https://api.probex.top

📖 Detailed Configuration Guide: See API Configuration Guide

🚀 Quick Start: See Quick Start Guide

Preferences

HyperPaper provides rich customization options:

AI Model Selection: Choose different AI models according to needs
Formula Processing Mode:
- No formula processing: Directly extract text without formula recognition
- Local OCR-based formula processing: Use local Pix2Text for OCR recognition, supports formula to LaTeX conversion
- VLM API-based formula processing: Use Vision API (such as Qwen-VL-Max) for recognition
Translation Target Language: Customize translation target language
Annotation Colors: Personalized annotation colors

🗺️ Roadmap

Completed ✅

Planned 🚧

🔈 Welcome to submit PRs and improve HyperPaper together!

🤝 Contributing

We welcome all forms of contributions!

How to Contribute

Fork this repository
Create your feature branch (git checkout -b feature/AmazingFeature)
Commit your changes (git commit -m 'Add some AmazingFeature')
Push to the branch (git push origin feature/AmazingFeature)
Open a Pull Request

Contributing Guidelines

Please see Contributing Guide for detailed contributing guidelines.

📄 License

This project is licensed under the MIT License. See the LICENSE file for details.

🙏 Acknowledgments

Qwen - Powerful large language model
Pix2Text - Excellent OCR tool

📮 Contact Us

GitHub Issues: Submit issues or suggestions
Pull Requests: Contribute code

⭐ If this project helps you, please give us a Star! ⭐

Name		Name	Last commit message	Last commit date
Latest commit History 15 Commits
HyperPaper		HyperPaper
Scripts		Scripts
docs		docs
.gitignore		.gitignore
README.md		README.md
VERSION		VERSION
version.sh		version.sh

Mengqi-Lei/HyperPaper

Folders and files

Latest commit

History

Repository files navigation