# datalab-to/marker

Convert PDF to markdown + JSON quickly with high accuracy

Repository: https://github.com/datalab-to/marker
Canonical: https://ross.abutalabs.com/products/marker
Homepage: https://www.datalab.to
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-08-26T21:36:48+00:00

## Health v2 (maintenance only)
Score: 92/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 93, longevity 74
- inputs: {"age_days": 1038, "days_push": 7, "days_rel": 44, "gap_med": 3.0, "n_releases_24m": 59}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 39295, forks 2826 (observed 2026-08-28T04:12:07.808461+00:00)

## What it is
Marker is a Python library and CLI tool that converts PDFs, images, and office documents (DOCX, PPTX, XLSX, EPUB, HTML) into markdown, JSON, HTML, and chunks with high accuracy. It uses deep learning models for layout detection and OCR, with optional LLM-boosted accuracy, and runs on GPU, CPU, or MPS.

## Use cases
- convert pdf to markdown
- extract text and tables from scanned pdfs
- turn documents into json for llm ingestion
- ocr pdfs with equations and math
- batch convert docx and pptx to markdown
- prepare documents for rag pipelines
- extract images from pdf documents

## When to choose
- you need fast, accurate pdf-to-markdown conversion at scale
- your documents contain tables, math, forms, or multi-column layouts
- you want a self-hosted alternative to managed document APIs
- you need to feed documents into LLM or RAG workflows

## When to avoid
- you only need simple text extraction from born-digital PDFs (pdftotext may suffice)
- you cannot run deep learning models locally and don't want the managed platform
- you need pixel-perfect visual fidelity rather than structured markdown output

## Facets
- artifact type: library
- maturity: active
- function: ocr, pdf, parser, machine-learning, llm-inference
- domain: pdf, files, machine-learning, developer-tools
- platform: python, cli, cross-platform
- tags: pdf-to-markdown, document-conversion, document-intelligence, deep-learning, ocr, gpu

## Member repositories
- datalab-to/marker (main) score 92

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:12:07.808461+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:24:10.905619+00:00, confidence not recorded.
  - readme: https://github.com/datalab-to/marker (fetched 2026-08-28T04:12:07.808461+00:00, sha 0f6465b71e0b)
  - homepage: https://www.datalab.to (fetched 2026-08-29T07:46:47.144810+00:00, sha 2bbd3c9c83e2)
- Data as of 2026-08-30T08:39:29.467469+00:00.