How GetMarkdown Works
GetMarkdown converts common document, web, text, image, and audio formats to clean, structured Markdown entirely in your browser. No files are uploaded to any server. The average conversion takes under 10 seconds for typical files.
How does browser-based file conversion work?
GetMarkdown uses Pyodide, a port of CPython compiled to WebAssembly, to run Python directly in your browser. When you drop a file, the browser loads a lightweight Python runtime (~6 MB compressed), installs the MarkItDown conversion library, writes your file to an in-memory virtual filesystem, and runs the conversion. The entire process happens on your device. According to Mozilla research, WebAssembly executes at near-native speed, typically within 2x of compiled C code, making complex document parsing practical in a browser tab.
What file formats does GetMarkdown support?
GetMarkdown supports common file formats across four categories. Each format is parsed with MarkItDown in the browser, preserving document structure such as headings, tables, lists, and embedded media when that structure is available.
Documents
- PDF (.pdf)
- Word (.docx)
- PowerPoint (.pptx)
- Excel (.xlsx)
Web & Text
- HTML (.html)
- HTML (.htm)
- Plain text (.txt)
Images
- JPEG (.jpg)
- JPEG (.jpeg)
- PNG (.png)
Audio
- MP3 (.mp3)
- WAV (.wav)
Is GetMarkdown private and secure?
Yes. In the free tier, zero bytes of your file data are transmitted to any server. The conversion runs entirely in a WebAssembly sandbox inside your browser tab. There are no cookies, no analytics on file content, and no server-side logging of what you convert. The 10 MB per-file size limit exists because browser memory is finite, not because of bandwidth. Over 100% of free-tier processing happens client-side. For organizations handling sensitive documents (legal, medical, financial), this architecture means compliance teams do not need to evaluate a third-party data processor.
How is the Markdown output structured?
GetMarkdown produces clean, standards-compliant CommonMark output. Document headings map to Markdown headings (#, ##, ###). Tables convert to pipe-delimited Markdown tables. Lists preserve nesting and ordering. Bold, italic, and inline code are retained. For DOCX and PPTX files, embedded images are extracted and packaged in a ZIP archive with correct relative paths in the Markdown. The output is designed to be immediately usable in any Markdown-based tool: Obsidian, Notion, GitHub, MkDocs, Docusaurus, or any LLM context window.
Can I use GetMarkdown output with AI agents and LLMs?
Yes. Markdown is the preferred input format for most large language models and RAG (Retrieval-Augmented Generation) pipelines. Unlike raw PDF or DOCX, Markdown preserves semantic structure (headings, lists, tables) without binary overhead, making it significantly more token-efficient for LLM processing. A typical 10-page Word document converts to roughly 2,000-4,000 tokens in Markdown, compared to noisy extracted text that can be 3-5x larger. Teams building AI agents, chatbots, and knowledge retrieval systems use GetMarkdown to preprocess documents before embedding or indexing.
What is the Pro PDF conversion?
The Pro tier handles PDFs that need better table extraction and complex layout handling. Unlike the free browser-based conversion, Pro uses server-side processing with specialized PDF parsing. It returns a ZIP containing both Markdown and structured JSON output. Pro is designed for PDFs with dense tables, multi-column layouts, or scanned content that benefits from advanced OCR. The free tier works well for most documents, but Pro delivers higher fidelity on challenging PDFs.
How does GetMarkdown compare to other converters?
Most file-to-Markdown converters require uploading your files to a server. GetMarkdown is one of the few tools that runs the full conversion pipeline in the browser. This means no file size quotas tied to server costs, no waiting in processing queues, and no privacy trade-offs. The underlying MarkItDown library (maintained by Microsoft) handles the actual parsing, so format support and output quality match what enterprise-grade tools produce. The main trade-off is the 10 MB file size limit imposed by browser memory constraints.
Can I convert multiple files at once?
Yes. GetMarkdown supports batch conversion of up to 8 files simultaneously. Drop multiple files onto the upload area and each one is converted independently. Results are available as individual Markdown downloads or a combined ZIP archive.