doc7
A CLI and MCP tool that turns PDFs, Office files, scans, screenshots, charts and diagrams into AI-ready Markdown, using any OpenAI-compatible multimodal model instead of a fixed OCR pipeline.
doc7 converts PDFs, Office files, scans, screenshots, charts, formulas, and diagrams into Markdown that an AI system can search, quote, and reason over, using any OpenAI-compatible multimodal model rather than a required OCR stack or a document-processing service. It discovers a local vision model, such as one running in LM Studio or Ollama, on first run, and saves the selection for later use; a remote endpoint can also be configured.
Rather than stopping at character-level text extraction, it reads a whole page, recovering figures, displayed equations, footnotes, and the relationships between diagram elements as searchable Markdown, and reassembles multi-page documents in order. In one published comparison run against two other document-conversion tools, using the same local model and endpoint, doc7 recovered more of a set of checked visual facts from two raster-only PDFs than either alternative.
It ships as a single Go binary that provides an interactive CLI, batch processing, model checks, an MCP server, a Go SDK, and an asynchronous HTTP service.