Workflowy offers a simpler way to stay organized. If you have a crazy job or an ambitious project, we will be your trusty sidekick.

TokenPig converts documents into clean, structured Markdown built specifically for LLM and RAG workflows — ChatGPT, Claude, Gemini, and retrieval pipelines.

The problem

Raw PDF, Word, PowerPoint and Excel exports carry a lot of formatting noise — repeated headers, broken tables, inconsistent whitespace — that eats into an LLM's context window without adding useful information. Cleaning that up manually is tedious, and general-purpose converters weren't built with token efficiency in mind.

What TokenPig does

• Converts PDF, DOCX, PPTX, XLSX, HTML, CSV, JSON, XML, TXT and MD into clean Markdown
• Estimates how many tokens each conversion saves versus the raw document
• Runs entirely in the browser — no installation, no code, no configuration
• Offers three output modes (Clean, Compact, Max Savings) depending on how aggressively you want to strip formatting

Plans

• Free — try it with no signup
• Personal — for individuals doing regular one-off conversions
• Pro / Enterprise — batch processing, ZIP export, and a conversion API for teams automating document ingestion into RAG pipelines

Who it's for

Researchers, consultants, and students who regularly paste documents into ChatGPT or Claude and want cleaner, cheaper context — plus developers and teams building retrieval-augmented generation pipelines who need reliable document-to-Markdown conversion, with an API for automation.

Compared to open-source libraries like MarkItDown or Docling, TokenPig is built for people who want a ready-to-use tool with no setup, along with visibility into token savings — while still offering an API for teams that want to integrate it into their own pipeline.