Word, Excel & PowerPoint to Markdown — LLM-Ready Text & JSON

Convert Word (.docx), PowerPoint (.pptx), Excel (.xlsx), OpenDocument (.odt/.ods), EPUB and CSV URLs to clean, LLM-ready Markdown, plain text and typed JSON: headings, lists and tables kept, every sheet as rows, slides with speaker notes, RAG chunks and metadata. $3 per 1,000 documents.

Turn Word, Excel and PowerPoint files into clean, LLM-ready Markdown and JSON. Paste document links (or a web page that links to documents) and get one row per file. Word (.docx) keeps its headings, bullet and numbered lists, tables, links and footnotes; OpenDocument text (.odt) keeps headings, lists, tables and links. PowerPoint (.pptx) becomes one section per slide with the slide title, bullets, tables and speaker notes. Excel (.xlsx), OpenDocument spreadsheets (.ods) and CSV/TSV come back as a Markdown table per sheet plus typed JSON rows keyed by column name, with numbers, booleans, formula results and real ISO dates instead of Excel serial numbers. EPUB e-books are read chapter by chapter in reading order. Each row also has title, author, subject, keywords, language, created and modified dates, page/slide/sheet counts and word count, plus optional plain text and ready-to-embed RAG chunks that carry their heading path. Use it to feed RAG pipelines and vector databases, build chatbot knowledge bases from policies, manuals and slide decks, give AI agents a document-reading tool, or turn government and corporate spreadsheets into JSON. Simple pricing: $3 per 1,000 documents, every sheet, slide and format included, with no compute bill on top. Failed downloads, unsupported or corrupt files and duplicates are never charged.

What data does it return?

Every run outputs clean, schema-validated rows like these:

url	format	title	wordCount	markdown	sheetCount	sheets	slideCount
https://calibre-ebook.com/downloads/demos/demo.docx	docx	DOCX Demo	1588	# Demonstration of DOCX support in calibre\n\n…\n\n| ITEM | NEEDED |\n| --- | --- |\n| Books | 1 |
https://example.org/q3-results.xlsx	xlsx				3	[{ name: "Revenue", headers: ["Region", "Q3", "Updated"], rows: [{ Region: "EMEA", Q3: 1250000, Updated: "2026-09-30" }] }]
https://example.org/board-deck.pptx	pptx			## Slide 1: Q3 Board Update\n\n- Revenue up 12%\n\nSpeaker notes:\n\nOpen with the headline number…			24

How much does it cost?

Transparent pay-per-event pricing — you only pay for what you get, and the actor always respects your run budget: set a maximum total charge and it stops cleanly at the cap instead of overrunning.

Event	Price	Charged	Worked example
document-converted	$0.003	per document converted and delivered (all sheets, slides, output formats and chunks included)	1,000 documents = $3.00; a 30-sheet workbook is still $0.003
apify-actor-start	$0.00005	per run per GB of memory (platform start fee)	One run at 1 GB = $0.00005

Input options

Field	Type	Required	Default	Description
urls	array	yes	—	Links to Word (.docx), PowerPoint (.pptx), Excel (.xlsx), OpenDocument (.odt, .ods), EPUB, CSV/TSV, Markdown or text files. You can also add a web page (e.g. a statistics or publications page): every document it links to is converted.
outputFormats	array	no	["markdown"]	Which content fields to return. Same price whatever you pick.
spreadsheetJson	boolean	no	true	For Excel, OpenDocument and CSV files, also return every sheet as typed JSON rows keyed by column name (numbers, booleans and ISO dates, not just text).
headerRow	boolean	no	true	Use each sheet's first non-empty row as column names. Turn off to name columns A, B, C…
maxRowsPerSheet	integer	no	10000	Read at most this many data rows from each sheet. Longer sheets are marked truncated.
includeHiddenSheets	boolean	no	false	Also convert Excel sheets that are hidden in the workbook.
includeSpeakerNotes	boolean	no	true	Add each PowerPoint slide's speaker notes after its content.
chunkSize	integer	no	0	Also split each document into chunks of at most this many characters for embedding, split at headings, paragraphs and table rows, each with its heading path. 0 = no chunks. Around 1,500–4,000 suits most embedding models.
maxDocuments	integer	no	1000	Stop after converting this many documents (including ones found on web pages).
findDocumentLinks	boolean	no	true	When a URL is a web page rather than a document, convert the documents it links to (one level, robots.txt respected).

Limits & honest notes

• Supported: .docx/.docm, .pptx/.pptm, .xlsx/.xlsm, .odt, .ods, .epub, .csv, .tsv, .md and .txt. Legacy binary files (.doc, .xls, .ppt), .odp, RTF and password-protected documents are reported and not charged. PDFs have their own actor: PDF to Markdown.
• Images, charts and SmartArt are skipped (they have no text or URL to keep). Bold and italic survive in Word files; colours, fonts and layout do not.
• Excel formulas return the value saved in the file (the last calculated result); they are not recalculated. Merged cells keep their value in the first cell only.
• CSV values stay as text so postcodes, IDs and phone numbers keep their leading zeros. Excel and OpenDocument cells are typed.
• Files up to 100 MB. Spreadsheets return up to 100,000 rows per sheet (10,000 by default). Very large results move to a linked file, with a preview kept in the row.
• Web pages in the input are scanned for links to documents (one level deep, robots.txt respected). If a site's robots.txt can't be fetched (HTTP 5xx or a network error, after one retry), the documents linked on that site are skipped (RFC 9309) and the page gets a free row saying why. Pages behind a login are not supported.
• Respects your run budget: set a maximum total charge and conversion stops cleanly at the cap, with no document charged twice.

Related ActorForge actors

• Word to Markdown
• Excel to JSON
• PDF to Markdown — the same output for PDFs
• Website to Markdown

Changelog

• 2026-10-18 — Initial release: Word, PowerPoint, Excel, OpenDocument text and spreadsheets, EPUB and CSV to Markdown, plain text and typed JSON rows; speaker notes, metadata, RAG chunks, document-link discovery on web pages.

---

Built and maintained by ActorForge. Reliability first: every actor in the portfolio is monitored by an automated nightly self-test, and breakages are typically fixed the same day.