JustMarkdown

JustMarkdown is a "document-to-Markdown workspace." Its core idea is: conversion is only the first step; the conversion result is retained as editable Markdown, allowing users to continue reading, annotating, asking AI questions, and exporting within the same space, without having to repeatedly copy and paste between a reader, an AI assistant, and a note-taking app.

Main capabilities

• Document conversion: convert sources such as PDF, Word, Excel, EPUB, HTML, images, and audio into Markdown, extracting headings, tables, code, and OCR text, and removing layout noise.
• Broad format coverage: the official statement says it provides 18 input formats and 31 tools, with a focused page for each tool.
• Continue processing within the workspace: after conversion, you can edit, annotate, and reorganize content in the same space, ask AI questions about the document, and export to PDF, Word, HTML, or images.
• MCP server: provides a hosted /api/mcp endpoint that can be used in Claude, Cursor, or any MCP client; it also provides the local stdio bridge command npx -y justmarkdown.

Conversion notes for each input format

• PDF: uses MinerU to extract text, tables, and code blocks, and outputs clean Markdown, suitable for researchers, archivists, and anyone who needs editable content from PDFs.
• Word .docx: converts to structured Markdown, preserving headings, tables, bold formatting, and nested lists, suitable for migrating documents, blog drafts, and manuscripts.
• PPTX: processed locally in the browser; the official statement says it is fast, free, and completely private, and outputs AI-friendly Markdown.
• Excel .xlsx/.xls: converts each worksheet into a well-formatted Markdown table, suitable for embedding in documents, READMEs, and knowledge bases.
• Images: free online image-to-Markdown; upload screenshots, photos, or scans and use OCR to extract text as Markdown.
• Web URL: extracts the main content of a web page and removes clutter, suitable for LLM training, RAG, and archiving.
• HTML files: removes noisy markup, scripts, and styles, and extracts clean, readable Markdown, suitable for archiving web pages, migrating content, and cleaning up exported documents.
• EPUB: converts in chapter order to Markdown that preserves the reading flow, suitable for extracting reading notes, study materials, and migrationLong-form content.
• Audio: Convert meeting recordings, interviews, podcasts, and voice notes in MP3, WAV, M4A, and FLAC into searchable Markdown; the official statement says transcription uses its own speech recognition engine, is billed per minute, and includes punctuation and paragraph segmentation.
• CSV/TSV: Convert into well-aligned Markdown tables, making them easy to embed in READMEs, documentation, and GitHub wikis.
• JSON: Wrap as syntax-highlighted Markdown code blocks, suitable for API documentation, technical specifications, and structured data display.
• YAML: Render as Markdown code blocks that preserve indentation, suitable for documenting API specifications, CI/CD configurations, and static site frontmatter.
• Source code: Package as automatically detected, syntax-highlighted Markdown code blocks, supporting more than 15 languages including TypeScript, Python, Go, Rust, and Java.
• Jupyter .ipynb: Convert to Markdown in order, preserving code cells, outputs, and markdown cells, suitable for documenting data science workflows, sharing research, and publishing technical tutorials.

How to use

1. Open the workspace, drag files in, or enter a web URL.
2. After getting clean Markdown, continue reading, annotating, editing, and asking AI questions within the space.
3. Export as PDF, Word, HTML, or images within the same space. The official statement says there are no plugins, no batch processing queue, and no installation required. Also, in MCP clients, you can add an endpoint with a configuration block, with no local installation package required.

MCP server capabilities The official statement lists that the hosted /api/mcp endpoint provides six tools: convert_url and convert_base64 convert documents and URLs into clean Markdown; request_upload_url and convert_upload handle files that are too large to use base64 (requires an API Key); list_supported_formats and usage_status report supported formats and quota. This endpoint can be used anonymously with a small daily quota; adding an API Key unlocks OCR, larger uploads, and higher limits. Clients that can only start local processes can run npx -y justmarkdown, which is a stdio bridge to the same endpointwith no key required. API Keys are created in account settings and are optional.

Use cases The official statement gives three usage patterns: parse technical documentation into code-friendly Markdown; summarize a 50-page paper into a literature review; and provide AI agents with a truly usable documentation tool. Suitable for creators, researchers, and knowledge workers to turn messy source material into clearer drafts, reports, notes, or published documents.

Limitations and notes

• The official statement says anonymous use of the remote endpoint has a small daily quota, and higher limits and OCR require an API Key.
• Audio transcription is billed per minute; specific pricing information is not provided.
• PPTX conversion is processed locally in the browser.
• The website homepage says you can try it without registration, and API Keys are optional and created in account settings.
• Other specific charges, prices, and platform details are not yet provided.