Sygal is a desktop tool that turns documents into clean, AI-ready Markdown and structured JSON directly on your machine. It handles PDFs, Word and Excel files, images, scans, web pages and even YouTube links, converting them locally so they can be dropped into a prompt, a knowledge base or a data pipeline.
The tool is especially useful for scans, images and web pages. Those formats carry pixels, layout and markup that an AI model pays to read but never uses. Once converted to text, the same content can run roughly 2 to 7x lighter on tokens, so the model reads what matters. YouTube links arrive as clean transcripts, ready to summarize or search. Already-structured files such as Word, Excel and PowerPoint are standardized into the same consistent output.
The workflow is simple: drop a file, a URL or a screenshot pasted with Cmd+V, convert it in about 1.7 seconds per file, review the original alongside the result, then export as Markdown, JSON or both. Converted files stay in a searchable local library, and each document shows its token count. Sygal supports 29 formats and lets users bring their own OCR provider, such as OpenAI, Claude, Mistral, NVIDIA or Gemini, using their own API key.
Sygal is aimed at developers building RAG pipelines and agents, at legal, accounting, HR and healthcare professionals who need sensitive documents kept on their machine, and at everyday users managing notes and research. It is offered as a €25 one-time purchase with no subscription and no account, with a 14-day free trial on macOS 12+ and Windows 10+.
Source: Hacker News · Summarized by HeadlinesBriefing