Pricing

from $2.40 / 1,000 documents

Try for free

Go to Apify Store

PDF to Markdown Extractor & Document Parser: Tables, JSON, OCR

Try for free

Extract text and tables from PDF, Word, Excel, PowerPoint, HTML and scanned images. Returns LLM-ready Markdown, tables as JSON/CSV, RAG chunks and metadata. OCR for scanned PDFs, optional AI field extraction (invoices, forms). No API key; pay per document.

Pricing

from $2.40 / 1,000 documents

Rating

0.0

(0)

Developer

Fernando Guiraud

Actor stats

Bookmarked

Total users

Monthly active users

7 hours ago

Last modified

Categories

Developer tools

Automation

Document URLs

sources

Optional

Document URLs. Supported: PDF (native or scanned), DOCX, XLSX, XLS, PPTX, HTML pages, CSV, TXT and images. Share links from Google Drive, Google Docs/Sheets/Slides, Dropbox, OneDrive/SharePoint and GitHub are converted to downloads automatically (the file must be shared as 'Anyone with the link'). The real file type is detected from the content.

Type:array

Files as base64 (optional)

base64Files

Optional

For files that are not reachable by URL, e.g. from an AI agent or a local script: a list of objects {"fileName": "report.pdf", "content": ""}. A data: URI is also accepted as content. Keep the total input under ~9 MB; use URLs for bigger files.

Type:array

Also process documents linked from web pages

followDocumentLinks

Optional

If a source URL is a web page, find links to PDF/DOCX/XLSX/XLS/PPTX/CSV files on it and convert those documents instead (e.g. a reports, filings or publications page). The page itself is not billed.

Type:boolean

Default:false

Max linked documents per page

maxLinkedDocuments

Optional

Limit for 'Also process documents linked from web pages'.

Type:integer

Minimum:1

Maximum:500

Default:20

What to return

outputs

Optional

Fields to include for each document. 'markdown': whole document as LLM-ready Markdown (headings, lists, tables). 'tables': every table as header + rows + CSV. 'text': plain text. 'pages': Markdown per page/slide/sheet. 'chunks': Markdown split into RAG chunks with page numbers.

Type:string[]

Default:

[
  "markdown",
  "tables"
]

Also save downloadable files

saveFiles

Optional

Save each document's Markdown as a .md file, each table as a .csv file and all tables together as an Excel workbook (.xlsx, one sheet per table) in the run's key-value store; the result includes their download links (handy without code: open the link, get the file).

Type:boolean

Default:false

PDF table detection

tableDetection

Optional

'lines': detect tables drawn with borders/grid lines (precise, recommended). 'text': also detect borderless tables by aligning text columns (experimental: finds more tables, but columns can be split imperfectly).

Type:string

Default:lines

Options:

linestext

Remove running headers and footers

removeHeadersFooters

Optional

Drop lines repeated at the top or bottom of most PDF pages (page numbers, document titles, confidentiality notices) so Markdown and RAG chunks stay clean.

Type:boolean

Default:true

OCR mode

ocr

Optional

'auto': run OCR only on scanned pages that contain no text (recommended). 'always': OCR every page, even if it has a text layer. 'never': no OCR (images and scanned pages return no text). OCR pages are billed separately.

Type:string

Default:auto

Options:

autoalwaysnever

OCR languages

ocrLanguages

Optional

Languages of the scanned text. Adding more languages is slightly slower.

Type:string[]

Default:

[
  "eng"
]

Page range

pageRange

Optional

Pages to process, e.g. '1-5, 8'. Applies to PDF pages, PowerPoint slides and Excel sheets. Leave empty for all.

Type:string

Max pages per document

maxPages

Optional

Safety limit on pages processed per document.

Type:integer

Minimum:1

Maximum:5000

Default:500

Chunk size (characters)

chunkSize

Optional

Maximum chunk length when 'RAG chunks' output is selected. Chunks follow paragraph boundaries.

Type:integer

Minimum:100

Maximum:20000

Default:1000

Chunk overlap (characters)

chunkOverlap

Optional

Characters from the end of the previous chunk repeated at the start of the next one, to keep context.

Type:integer

Minimum:0

Maximum:5000

Default:100

Max file size (MB)

maxFileSizeMb

Optional

Documents larger than this are skipped (not billed).

Type:integer

Minimum:1

Maximum:500

Default:50

PDF password

pdfPassword

Optional

Password for encrypted PDFs (applies to all PDFs in the run).

Type:string

Max characters per document

maxCharactersPerDocument

Optional

Truncate 'markdown' and 'text' to this many characters (0 = no limit). Useful for AI agents with limited context windows.

Type:integer

Minimum:0

Maximum:10000000

Default:0

Fields to extract (AI, optional)

extractionFields

Optional

Pull specific values out of every document with Claude, e.g. ["invoice_number", "invoice_date", {"name": "total", "type": "number"}, "vendor_name"]. Results appear in 'extracted'. Requires your Anthropic API key. Billed as one extra 'AI extraction' event per document (Claude usage is billed to your own Anthropic account).

Type:array

Extraction JSON Schema (advanced)

extractionSchema

Optional

Instead of a field list, a full JSON Schema (type: object) for nested data, e.g. invoice line items. Every property is returned; missing values come back as null.

Type:object

Extraction instructions

extractionInstructions

Optional

Optional guidance, e.g. 'Dates as YYYY-MM-DD; amounts without currency symbols'.

Type:string

Anthropic API key

anthropicApiKey

Optional

Your key from console.anthropic.com. Stored as a secret input; used only to call Claude for this run.

Type:string

Claude model

extractionModel

Optional

'claude-opus-5': most accurate (default). 'claude-sonnet-5': cheaper, great for invoices/forms. 'claude-haiku-4-5': cheapest, for simple fields.

Type:string

Default:claude-opus-5

Options:

claude-opus-5claude-sonnet-5claude-haiku-4-5

Fail the run if any document fails

failOnError

Optional

Mark the run as FAILED when a document cannot be processed or yields no text. Useful for pipelines and monitoring. Successfully processed documents are still returned (and billed).

Type:boolean

Default:false

Max parallel documents

maxConcurrency

Optional

How many documents are processed at the same time.

Type:integer

Minimum:1

Maximum:8

Default:2

OCR & Document Extractor – PDF & Image to Text, JSON, Word

lofomachines/ocr-document-extractor

Convert scanned PDFs and images into clean, structured text in bulk. Export to JSON, Markdown, DOCX, TXT or HTML with tables and layout preserved.

Lofomachines

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

PDF to Markdown Converter - AI-Powered with OCR & Tables

clearpath/pdf-to-markdown-api

Convert PDFs to clean Markdown with GPU-accelerated AI. Extracts tables, LaTeX formulas, and images from complex layouts. Supports OCR for scanned docs in 8 languages. Batch process hundreds of PDFs in parallel via URL, upload, or API.

ClearPath

PDF Text Extractor & Document Parser: DOCX, XLSX, OCR, Markdown

yabanana99/pdf-word-excel-to-markdown

Parse Word, PowerPoint, Excel and PDF documents into clean Markdown and structured JSON: tables extracted cell-by-cell, document metadata, and RAG-ready chunks with heading paths. Built-in OCR for scans and images.

Daniele Giovane

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

PDF Extractor 2.0

jupri/pdf-extractor-2-0

💫 Extract PDF Document Contents including Metadata, Images, Pages, Tables, Attachments, etc.

cat

173

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.

Scrape Sage

Document Text Extractor

lergassy/document-text-extractor

Extract text, Markdown, tables and metadata from PDF, Word, Excel, PowerPoint, EPUB, HTML and CSV files. Reads scanned pages with OCR in 19 languages and splits documents into RAG-ready chunks. Export data, run via API, schedule runs, or integrate with other tools or AI workflows.

Matvey

PDF to Markdown RAG-Ready

hedelka/pdf-to-markdown-rag

Premium PDF scraper that preserves tables and structure. Optimized for RAG.

Dmitry Goncharov

PDF Text Extractor — Text, Markdown, Tables and OCR

lergassy/pdf-text-extractor

[$0.003/PDF] Extract text from PDF files by URL, upload or base64: full text, per-page text, Markdown, tables and metadata. Scanned pages are read with OCR in 19 languages. Password PDFs, page ranges, RAG chunks. No fee per run.

Matvey