Canonizr

Canonizr converts a wide range of document types—including PDFs, DOCX files, and scanned images—into markdown formatted for language‑model consumption. It extracts text, headings, tables, and other structural elements while preserving the original layout, delivering clean, machine‑readable output that can be fed directly into downstream processing pipelines.

The tool is intended for developers building agents or applications that need to ingest and reason over user‑provided documents without storing the raw content. By operating as a self‑hosted service or through an API, it allows integration into private infrastructures while guaranteeing that no data is retained after processing.

Its distinguishing characteristic is the combination of format‑agnostic parsing with a strict zero‑retention policy, ensuring that sensitive documents never leave the host environment. This makes it suitable for privacy‑focused workflows where compliance and data minimization are required.