Pricing

from $3.00 / 1,000 pdf converteds

Try for free

Go to Apify Store

PDF to Markdown API & Converter: PDF to Text for AI Agents, RAG

Try for free

PDF to markdown converter and MCP tool for AI agents and RAG: give PDF URLs, get clean Markdown with headings, page texts, title, author and word count. Text-layer PDFs only, no OCR. Pay $0.003 per PDF converted (up to 30 pages), $0.0001 per extra page, plus a small start fee per run.

Pricing

from $3.00 / 1,000 pdf converteds

Rating

0.0

(0)

Developer

PlainRows

Actor stats

Bookmarked

Total users

Monthly active users

5 days ago

Last modified

Typical uses

RAG and knowledge bases: feed research papers, manuals, reports and documentation into a vector store as Markdown, page by page.
AI agents (MCP): let an agent read a PDF it found on the web in one tool call.
Summaries and extraction: give an LLM the Markdown of contracts, filings, datasheets or whitepapers.
Search and archiving: index the full text of public PDFs with their page count and word count.

What you get for each PDF

Field	Description
`url`, `finalUrl`	The PDF address you gave, and the address after redirects
`found`	`true` when the PDF was converted (billed row); `false` rows are free and say why in `error`
`title`, `author`	From the PDF metadata (`null` when the file has none: nothing is guessed)
`pageCount`, `pagesConverted`	Pages in the file, and pages converted (at most `maxPagesPerPdf`)
`markdown`	Markdown of the converted pages, with a `<!-- page N -->` marker before each page
`pages`	With `includePageTexts`: plain text of each page, `[{ "page": 1, "text": "..." }]`
`wordCount`, `charCount`	Words in the text and characters in the Markdown (useful to size LLM context)
`error`, `errorCode`	Why a PDF was not converted (`NOT_A_PDF`, `TOO_LARGE`, `HTTP_ERROR`, `NO_TEXT_LAYER`, `UNREADABLE_PDF`, `ENCRYPTED_PDF`, `TIMEOUT`...), or a note when only part of it was converted
`convertedAt`	When the PDF was processed (UTC)

Headings are approximated from font size (a line clearly larger than the body text becomes #, ## or ###). Tables, images and multi-column layouts come out as plain text in reading order as stored in the PDF; words split by a hyphen at a line end are joined.

Input

{
  "urls": ["https://arxiv.org/pdf/1706.03762", "https://www.irs.gov/pub/irs-pdf/fw9.pdf"],
  "maxPagesPerPdf": 30,
  "includePageTexts": false
}

urls: public PDF addresses, up to 5,000 per run (duplicates removed). Files up to 50 MB.
maxPagesPerPdf (default 30, max 300): only the first pages up to this number are converted.
includePageTexts (default false): also return the text of each page.

Output example

{
  "url": "https://arxiv.org/pdf/1706.03762",
  "found": true,
  "title": null,
  "author": null,
  "pageCount": 15,
  "pagesConverted": 15,
  "markdown": "<!-- page 1 -->\n\nProvided proper attribution is provided, Google hereby grants permission...\n\n## Attention Is All You Need\n\nAshish Vaswani∗ Google Brain avaswani@google.com\n\n...\n\n### Abstract\n\nThe dominant sequence transduction models are based on complex recurrent or convolutional neural networks...",
  "pages": null,
  "wordCount": 6135,
  "charCount": 40118,
  "error": null,
  "errorCode": null,
  "finalUrl": "https://arxiv.org/pdf/1706.03762",
  "convertedAt": "2026-10-23T14:00:00.000Z",
  "schemaVersion": "1.0"
}

(Excerpt; this arXiv file has no title in its metadata, so title is null.)

Pricing

Pay per PDF converted, no subscription:

$0.003 per PDF converted (pdf-converted), up to 30 pages included;
$0.0001 per extra page (extra-page) beyond 30, only if you raise maxPagesPerPdf and the PDF is longer;
a small start fee per run (apify-actor-start, $0.002 per GB of memory).

Free rows: PDFs without a text layer (scans), unreadable or password-protected files, files over 50 MB, non-PDF URLs (HTML pages, images), and URLs that fail (404, timeout, invalid or private address). The run never spends more than the maximum cost you set; if a run keeps hitting free failures, it stops early and says so.

Free Apify plan: each run converts up to 10 PDFs, so you can check the output. A paid Apify plan removes this limit.

Use it from code, automations or AI agents

Call it from the Apify API, the Apify MCP server (as a tool for Claude, ChatGPT or any MCP client), Make, Zapier, n8n or LangChain. Results are in the run's dataset (JSON, CSV, Excel); the run summary is in the OUTPUT record of its key-value store.

FAQ

Does it do OCR? No. It reads the text layer that most digital PDFs contain (exported from Word, LaTeX, browsers...). A scanned PDF without text returns found: false, errorCode: NO_TEXT_LAYER, free.

Are the headings exact? They are an approximation from font size; the text itself is exactly what the PDF contains.

What about very long PDFs? Set maxPagesPerPdf up to 300. Each PDF also has a time limit; if it is reached, the pages already converted are returned and only those are charged.

Which PDFs can it fetch? Public http(s) URLs that return a PDF file. Private network addresses are refused. The Actor identifies itself with a PlainRows user agent.

More PlainRows tools

Same pay-per-result pricing, same clean JSON, built to chain together in n8n, Make, Zapier or AI agents.

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

Privacy Policy Risk Scanner - GDPR and CCPA Evidence API

changeable_peddler/privacy-policy-risk-scanner

Scan public website privacy pages for GDPR, CCPA, DSAR, retention, transfer, DPO, and cookie policy evidence with source-linked snippets for manual review.

Sean J

Website to Markdown — Clean Page Text for LLMs

sourcedirect/website-to-markdown

Fetch any URL and return clean markdown/text plus links. No headless browser for most pages. Free demo: https://vaytric.com/page-to-markdown/ — bulk URL lists here.

Kurt Landman

YouTube Transcript Scraper: Captions to Text, Timestamps & SRT

poised_beardlichen/youtube-transcript-scraper

Use this when you need YouTube video transcripts as text, timestamped segments or SRT. $0.004/transcript; videos without captions are free. Input: {"videos":["https://youtu.be/aircAruvnKk"],"languages":["en"]}. Output: text, segments, title, channel, language.

Isaac

Website to RAG Chunks — RAG crawler, token-sized, cited

webprobe-tools/website-to-rag-chunks

Crawl a docs site, help center or blog and get embedding-ready chunks: main content as Markdown, split at headings to a token size you choose, each with its heading path, a citation URL to its section, a token count and a content hash. Plain HTTP. Built and maintained by an AI agent.

WebProbe Tools

Technical Docs To Rag Markdown Chunker

arezdev/technical-docs-to-rag-markdown-chunker

Convert public documentation into clean Markdown and RAG chunks with source URLs, section context and content hashes. Discover pages through sitemaps and llms.txt, preserve code and tables, and export JSON for embedding and vector database pipelines.

Arezdev

Web Search to Markdown for RAG: Top Pages as Clean Text

tanod/web-search-to-markdown

Search the web and get up to 3 pages per query as Markdown for RAG, LLMs and AI agents. Up to 50 queries per run. USD 0.012 per query with readable pages on every Apify plan, plus the Actor start fee. No query charge for failed searches or queries with no readable page.

Tanod Labs

PDF to Text and OCR: Extract Text from PDFs and Scans

boubap/pdf-text-ocr

Extract the text of PDF files and scanned images, one row per page. Digital pages are read from their text layer at $0.50 per 1,000 pages, scans go through Tesseract OCR at $4 per 1,000 pages in 7 languages. Failed and blank pages are free.

Baptiste Bouton

Doc2RAG AI - Documentation to LLM-Ready Markdown Converter

automation_studio/doc-to-rag-markdown-converter

Convert technical documentation websites (Stripe, Supabase, Next.js, Mintlify, Docusaurus) into clean, LLM-ready Markdown with preserved multi-language code blocks, tables, and RAG frontmatter. Exports JSON dataset, single llms-full.txt, and ZIP archive.

Automation Studio

llms.txt Generator (Create llms.txt for Any Website)

pagecraft/llms-txt-generator

Generate a ready-to-upload llms.txt (and optional llms-full.txt) for any website from its sitemap and key pages, so ChatGPT, Claude and Perplexity understand and cite it. Bulk domains, no browser.