Export beyond Markdown

Clean Markdown is the core, but the result card’s single “Format…” menu turns it into whatever your workflow needs — pick a format, then hit Download (it never fires on its own). Powered by Pandoc, with no LLM involved.

Family	Formats
Markdown	.md, compact (whitespace-stripped), RAG chunks (.jsonl)
Office & ebook	Word .docx, ODT, EPUB
Web & typesetting	HTML, LaTeX, reStructuredText
Structured XML	DocBook, JATS, TEI, OPML
Data	JSON, YAML, TOON (the document as structured data)
Privacy	redacted PDF (PII blacked out — see Anonymization)

The Format menu

Everything lives in one place. Choose a target format from the dropdown, then press Download. Selecting a format never triggers a download by itself — you stay in control of when the file is produced.

Data formats (JSON · YAML · TOON)

Beyond documents, Escriba can hand you the result as structured data — the title, the sections (split by heading) and metadata (word/character counts), plus the raw Markdown. Pick JSON, YAML or TOON:

• JSON / YAML — the universal interchange formats, for pipelines and tooling.
• TOON (Token-Oriented Object Notation) — a compact, token-efficient representation designed for feeding data to LLMs: the sections collapse into a tidy tabular block, so you spend fewer tokens than JSON.

No AI is involved — it’s a local transformation of your document.

Compact Markdown

A whitespace-stripped variant of the Markdown that saves tokens without changing the content — handy right before pasting into a model with a tight budget.

These exports are pure conversions: no AI is involved, and nothing leaves your server.