TextLexiq is your smart document companion. Scan images or papers, extract text with OCR, and convert to PDF, Word, or LaTeX. Summarize, ask questions, and get insights; all powered by on-device AI for speed, privacy, and precision. Transform documents into knowledge with ease.
- Smart Scanning: Capture documents with edge detection, auto-cropping, and perspective correction.
- Advanced OCR: Extract text with high accuracy using on-device ML Kit or robust Tesseract (via Python/MCP).
- AI-Powered Insights: Summarize, simplify, and chat with your documents using local LLMs (Llama.cpp) or cloud models.
- Multi-Format Export: Convert your documents to professional PDF, DOCX (Word), or LaTeX formats.
- Privacy First: All processing can happen locally on your device or within your private network.
- Agent Ready: Exposes functionality as an MCP (Model Context Protocol) server for AI agents.
- Language: Kotlin
- UI: Jetpack Compose
- Architecture: MVVM
- Core Libraries: CameraX, ML Kit, OpenCV (Android), Room, Llama.cpp (Android)
- Language: Python
- Core Libraries: OpenCV (Headless), NumPy, PyTesseract, Pillow
- Purpose: Advanced image processing and OCR pipeline, mirroring the Android implementation for server-side/agent use.
- Language: TypeScript
- Framework: LeanMCP / Node.js
- Purpose: Exposes OCR capabilities as AI tools for agents (e.g., in Daytona or simple MCP clients).
- Open the project in Android Studio.
- Sync Gradle files.
- Run on an Android device or emulator (Camera required).
- Navigate to
packages/python-logic. - Install dependencies:
pip install -r requirements.txt. - Run OCR:
python main.py ocr <path_to_image>.
- Navigate to
packages/mcp-server. - Install dependencies:
npm install. - Build:
npm run build. - Start:
npm start.
This project is licensed under the MIT License - see the LICENSE file for details.