Pricing

from $2.00 / 1,000 document converteds

Try for free

Go to Apify Store

PDF to Markdown & Document to Markdown: OCR, RAG Chunks

Try for free

PDF to Markdown and document to Markdown converter for LLMs, RAG and AI agents. PDF, scanned PDF (OCR), DOCX, PPTX, XLSX, EPUB and HTML to clean Markdown or plain text, with tables, headings and table-aware RAG chunks. $0.003 per document.

Pricing

from $2.00 / 1,000 document converteds

Rating

0.0

(0)

Developer

Tobby Tools

Actor stats

Bookmarked

Total users

Monthly active users

7 days ago

Last modified

Categories

Developer tools

Automation

Files or file URLs

files

Optional

PDF, Word (DOCX), PowerPoint (PPTX), Excel (XLSX/XLS/CSV), EPUB, HTML, TXT/Markdown or image files. Paste public URLs or upload files. Google Drive, Google Docs, Dropbox and GitHub share links are converted to direct downloads automatically.

Type:array

Default:

[
  "https://arxiv.org/pdf/1706.03762"
]

Document URLs (bulk)

startUrls

Optional

Optional list of document URLs, e.g. pasted from a spreadsheet or uploaded as a text file. Each URL's userData is copied to the output.

Type:array

Base64 files (API / n8n / Make / Zapier)

base64Files

Optional

Send file bytes directly instead of a URL: a list of objects like {"fileName": "invoice.pdf", "data": ""}. A data: URI is also accepted. Handy for n8n, Make, Zapier and API calls where the file is not public.

Type:array

Example:

[
  {
    "fileName": "hello.txt",
    "data": "SGVsbG8gd29ybGQ="
  }
]

Output format

outputFormat

Optional

Markdown keeps headings, lists and tables (best for LLMs and RAG). Plain text removes all formatting.

Type:string

Default:markdown

Options:

markdowntext

Chunk size (characters)

chunkSize

Optional

Split each document into chunks of about this many characters for embeddings/RAG. Chunks break at headings, paragraphs and sentences, and record their pages and section. Set 0 to disable chunking.

Type:integer

Minimum:0

Maximum:100000

Default:1500

Chunk overlap (characters)

chunkOverlap

Optional

How many characters consecutive chunks share (at most half the chunk size).

Type:integer

Minimum:0

Maximum:20000

Default:150

One dataset item per chunk

datasetItemPerChunk

Optional

Output one row per chunk (with document metadata) instead of one row per document. Handy for loading straight into vector databases (Pinecone, Qdrant, Weaviate, Chroma integrations).

Type:boolean

Default:false

OCR (text recognition)

ocrMode

Optional

Auto: OCR only pages without a usable text layer (scans, photos, broken fonts). Always: OCR every page. Off: never OCR. OCR pages are billed separately.

Type:string

Default:auto

Options:

autoforceoff

OCR languages

ocrLanguages

Optional

Languages printed in the scans. Pick all that apply (fewer is faster and more accurate).

Type:string[]

Default:

[
  "eng"
]

Max OCR pages per document

maxOcrPagesPerDocument

Optional

Safety limit for OCR cost/time per document. Pages beyond it are listed in warnings.

Type:integer

Minimum:0

Maximum:2000

Default:100

Also save files

saveFiles

Optional

Save converted files to the run's key-value store; download links appear in each item's 'files'. Word = editable DOCX, Excel = every extracted table as a sheet, searchable PDF = OCR text layer added to scans.

Type:string[]

Default:

[]

Include page-by-page text

includePages

Optional

Add a 'pages' array with the text of every page (PDF, PPTX slides, XLSX sheets).

Type:boolean

Default:true

Include tables as data

includeTables

Optional

Add a 'tables' array with each table's rows (JSON) and Markdown.

Type:boolean

Default:true

Table format

tableFormat

Optional

How tables appear in the document content and chunks. HTML keeps merged/spanning cells exactly; Markdown is simpler. Each item in tables[] always has rows, markdown and (when cells are merged) html.

Type:string

Default:markdown

Options:

markdownhtml

Include per-page quality flags

includeQuality

Optional

Adds a quality object to each page: source (text-layer/ocr), OCR confidence, table count and confidence, image coverage, figure count and needsVision (route the page to a vision model).

Type:boolean

Default:true

Include layout elements with bounding boxes

includeElements

Optional

Adds elements[] to each page: headings, paragraphs, list items and tables with their bounding box [x0, top, x1, bottom] in PDF points (top-left origin). Useful for citations and highlighting. Chunks always carry page boxes for PDFs.

Type:boolean

Default:false

Page range

pageRange

Optional

Only convert these pages, e.g. '1-5, 8, 12-'. Empty = all pages.

Type:string

PDF password

pdfPassword

Optional

Password for encrypted PDFs (used for every PDF in the run).

Type:string

Detect tables

extractTables

Optional

Detect ruled and borderless tables and output them as Markdown tables.

Type:boolean

Default:true

Remove running headers, footers and page numbers

removeHeadersFooters

Optional

Drop text repeated at the top/bottom of pages (page numbers, running titles).

Type:boolean

Default:true

Max pages per document

maxPagesPerDocument

Optional

Only the first N pages of each document are converted.

Type:integer

Minimum:1

Maximum:10000

Default:500

HTML: main content only

htmlMainContentOnly

Optional

For web pages, keep the article/main content and drop menus, sidebars, cookie banners and footers.

Type:boolean

Default:true

Max file size

maxFileSizeMb

Optional

Larger files are skipped with a FILE_TOO_LARGE error item (not charged).

Type:integer

Minimum:1

Maximum:500

Default:100

Max documents

maxDocuments

Optional

Stop after this many documents.

Type:integer

Minimum:1

Maximum:100000

Default:10000

Parallel documents

maxConcurrency

Optional

How many documents are processed at once. Default: 1 per GB of memory.

Type:integer

Minimum:1

Maximum:16

Download timeout

downloadTimeoutSecs

Optional

Per-file download timeout.

Type:integer

Minimum:5

Maximum:900

Default:120

Key-value store with files

keyValueStoreId

Optional

Convert files stored in this key-value store (ID or name).

Type:string

Keys in that store

keyValueStoreKeys

Optional

Record keys to convert. Empty = every record in the store.

Type:array

PDF & DOCX to Markdown — Document Extractor for LLM/RAG

fetchbase/document-to-markdown

Convert PDF and Word (DOCX) documents into clean Markdown, text, or JSON. Smart PDF paragraph reflow, page markers for RAG citations, full DOCX structure (headings, lists, tables), custom auth headers. No browser — parses in seconds. Charged per page processed — no startup fee.

Fetchbase

Universal Web to Markdown (Bulk & AI-Ready)

lentic_october/web-to-markdown-converter

Bulk convert any website URLs to clean Markdown for AI & LLMs. Universal scraper that removes ads, scripts, and clutter. Optimized for RAG, ChatGPT, Claude, and LangChain. Fast, async, and API-ready.

kalthireddy Abhishek

PDF to Markdown & JSON Converter (Docling)

actorzlab/docling-pdf-converter

Convert PDF documents to clean Markdown, structured JSON, and plain text using IBM's open-source Docling AI. Handles text PDFs and scanned documents (OCR), extracts tables and images. No external API key required — runs fully on-device.

Khalil Drissi

Document Parser — PDF/DOCX to Markdown & JSON for RAG

genuine_qa/document-parser

Convert PDF, DOCX, PPTX, XLSX, HTML and images into clean Markdown or JSON for RAG and LLM pipelines. Powered by IBM's open-source Docling.

Rahul Bhiwagade

Tech Docs to LLM-Ready Markdown

hedelka/tech-docs-scraper

Scrapes technical documentation sites (Docusaurus, GitBook, MkDocs, ReadTheDocs) and converts them to clean, structured Markdown for RAG pipelines, LLM training, and AI assistants. Automatically detects documentation framework and removes navigation elements.

Dmitry Goncharov

Web-to-Markdown Generator for AI & RAG Pipelines

profitstack/web-to-markdown-generator-for-ai-rag-pipelines

Convert any website into clean, heading-based chunking, LLM-ready Markdown for RAG and AI agents.

Manas Mantri

Document to Markdown Converter for RAG

ottomath/document-to-markdown

Convert PDF, DOCX, XLSX, PPTX, HTML and EPUB links to clean Markdown or RAG-ready chunks. No OCR, no browser. $4 per 1,000 documents plus page fees.

Ottomath

OCR Text Extractor - Images & Scanned PDFs to Text

scrapesage/ocr-text-extractor

Extract text from images (PNG, JPG, TIFF, WEBP, BMP, GIF) and scanned PDFs with Tesseract OCR in 28 languages. Per-page text, confidence scores, optional word boxes, automatic text-layer detection for born-digital PDFs. No API key, no browser. JSON, CSV, Excel.