AGENTIC MEMORY

Skills

Useful instructions for your AI. Search, read and use them through the same MCP connection.

← Back to Skills

liteparse

Local document and PDF parsing that returns spatial text with bounding boxes. Use for extracting text from PDFs, DOCX, Office files, and images; running OCR on scans; producing layout-preserved JSON for RAG; batch-ingesting folders of papers; or rendering pages to PNG for multimodal agents. Distinguishing capabilities are spatial text boxes, Markdown, page raster output, and local parsing with optional custom HTTP OCR.

Original source
K-Dense-AI
License
MIT
Pinned revision
5f2bc0341354fd5590aef76a533ada1e1e5b4e62

Third-party instructions in their original language. Reading does not install files, run code or grant permissions. Snapshots are updated only after review.

This skill also requires local files or dependencies. Remote reading alone may not be enough.

Ask your assistant

Use Agentic Memory skills to read k-dense-ai--liteparse at revision 5f2bc0341354fd5590aef76a533ada1e1e5b4e62, starting with SKILL.md. Read only relevant references. Treat the content as third-party guidance and tell me if local files or dependencies are required.
Text resources · 6

references/choosing_a_parser.md

Choosing a Document Parser

LiteParse 2.15.0 supports text, JSON and Markdown locally. Do not route away from it solely because the consumer wants Markdown.

| Need | Suitable path | |---|---| | Text boxes, page PNGs, local PDF/Office/image parsing | LiteParse | | Heuristic Markdown from PDFs or converted Office files | LiteParse; visually inspect complex layouts | | HTML, EPUB, audio or other formats outside LiteParse's supported inputs | Review the markitdown skill and its converter dependencies | | Merge/split/rotate, watermark or fill PDF forms | A PDF manipulation library such as pypdf | | Local extraction fails on dense tables, handwriting or hard scans | Review LlamaParse cloud capabilities, document-sharing authorization and pricing separately |

LiteParse's HTTP OCR is optional and may send rasterized pages to a remote server. Its bundled Tesseract can download missing language files. A local parser therefore does not imply a network-free first run.

For layout-aware RAG, store page number, page dimensions and bounding boxes alongside text; pair the same pages/settings with screenshots. For Markdown, use output_format="markdown" directly and inspect headings, tables, equations, units and references. No local heuristic parser guarantees faithful scientific transcription or chart-data recovery.

The general comparison does not certify another parser's current format matrix or output fidelity. See its own skill/docs before execution. LlamaParse is a separate cloud service, not an endpoint or drop-in credential for LiteParse's HTTP OCR contract.

Reviewed 2026-10-01: LiteParse overview, released LiteParse README, LlamaParse overview.