LiteParse

Use it when

• You must parse PDFs, Office documents, or scanned images to markdown or structured JSON without uploading them to any cloud service
• You need a cheap pre-check that tells you which pages of a document actually need OCR so you can route or cost-estimate before a full parse

What it solves

Not the fit when

• the hardest documents (dense tables, multi-column layouts, charts, handwriting, scans), for which the README points to the cloud-based LlamaParse
• Office-format input without LibreOffice installed
• audio or video transcription
• high-accuracy extraction of dense tables and multi-column layouts
• handwriting recognition quality
• semantic document understanding with an LLM

Install

pip install liteparse

Invoke

lit parse document.pdf --format markdown -o output.md; check first whether OCR is needed with: lit is-complex document.pdf --quiet && lit parse document.pdf --no-ocr

Alternatives

No reviewed alternatives recorded yet.