noted.md is a command-line tool designed to convert handwritten notes from image or PDF formats into clean Markdown files. It leverages Large Language Models (LLMs) to transcribe text, recognize and format mathematical equations in LaTeX, and supports batch processing for multiple files. This tool is ideal for students, researchers, and anyone who prefers handwritten notes but needs them in a digital, editable format.

How It Works

The tool utilizes LLMs to perform Optical Character Recognition (OCR) and transcription on input files (PDF, JPG, JPEG, PNG). It intelligently identifies and formats mathematical equations using LaTeX syntax. Users can configure their preferred AI provider, choosing from Gemini, Claude, or Ollama, and manage credentials via an interactive setup or command-line flags.

Quick Start & Requirements

• Installation:
  • macOS/Linux (Recommended): brew tap tejas-raskar/noted.md && brew install notedmd
  • Windows: Download .zip from Releases, add bin to PATH.
  • Source: git clone https://github.com/tejas-raskar/noted.md.git && cd noted.md && cargo build --release
• Configuration: Use notedmd config --edit for interactive setup or flags like --set-provider and --set-api-key. Requires API keys for Gemini/Claude or a running Ollama instance.
• Usage: notedmd convert <path> to convert files. Supports --output directory, custom --prompt, and temporary API key overrides.
• Documentation: GitHub Repository

Highlighted Details

• Supports multiple AI providers: Gemini, Claude, Ollama, and OpenAI API compatible clients.
• Handles image (JPG, PNG) and PDF file formats.
• Automatic LaTeX formatting for mathematical equations.
• Batch processing capability for converting multiple files or entire directories.

Maintenance & Community

The project is hosted on GitHub and welcomes contributions via issues and pull requests.

Licensing & Compatibility

Licensed under the MIT License, permitting commercial use and integration with closed-source projects.

Limitations & Caveats

The accuracy of transcription and equation recognition is dependent on the underlying LLM provider and the quality of the input notes. Specific performance benchmarks are not provided.