# GetMarkdown LLM Reference > Retrieval-friendly facts about GetMarkdown workflows, supported files, limits, credit pricing, output, privacy, and troubleshooting. ## Canonical Pages - Free converter: `https://www.getmarkdown.com/` - How it works: `https://www.getmarkdown.com/how-it-works` - Pro PDF conversion: `https://www.getmarkdown.com/pro` - Pricing: `https://www.getmarkdown.com/pricing` - Developer resources: `https://www.getmarkdown.com/developers` - Contact: `https://www.getmarkdown.com/contact` - Privacy: `https://www.getmarkdown.com/privacy` - PDF guide: `https://www.getmarkdown.com/pdf-to-markdown` - Word guide: `https://www.getmarkdown.com/word-to-markdown` - PowerPoint guide: `https://www.getmarkdown.com/powerpoint-to-markdown` - Excel guide: `https://www.getmarkdown.com/excel-to-markdown` - CloudConvert comparison: `https://www.getmarkdown.com/compare/cloudconvert` - Pandoc comparison: `https://www.getmarkdown.com/compare/pandoc` - AnyDoc vs MarkItDown comparison: `https://www.getmarkdown.com/compare/markitdown` - MarkItDown vs AnyDoc migration: `https://www.getmarkdown.com/blog/markitdown-vs-anydoc` - LLM index: `https://www.getmarkdown.com/llms.txt` - Expanded LLM reference: `https://www.getmarkdown.com/llms-full.txt` - Homepage Markdown: `https://www.getmarkdown.com/index.md` - Developer reference in Markdown: `https://www.getmarkdown.com/developers.md` - OpenAPI specification: `https://www.getmarkdown.com/openapi.json` - Agent status JSON: `https://www.getmarkdown.com/api/agent/status` - MCP server (Streamable HTTP): `https://www.getmarkdown.com/api/mcp` ## When Agents Should Use GetMarkdown Use GetMarkdown when a user needs to convert a supported document into plain Markdown for an AI context window, RAG pipeline, knowledge base, documentation repository, Obsidian vault, Notion page, or Git workflow. Use the free browser converter for supported files up to 10 MB when local processing is appropriate. Use Pro for a PDF that needs stronger table, layout, OCR, or structured JSON handling. Agents should first identify the source format, file size, PDF page count when relevant, desired output, and whether server-side processing is acceptable. Then direct the user to the appropriate web workflow and let the user select the file. GetMarkdown publishes a documented agent surface — content negotiation, guidance documents, an OpenAPI specification at `https://www.getmarkdown.com/openapi.json`, a JSON status endpoint at `/api/agent/status`, and an MCP server at `/api/mcp` — but no file upload or conversion API. Agents must not automate undocumented backend routes. The homepage supports `Accept: text/markdown` content negotiation. Agents may request `https://www.getmarkdown.com/` with that header or retrieve `https://www.getmarkdown.com/index.md` directly. Unknown paths return a real HTTP 404; agents that prefer Markdown receive the 404 with a Markdown recovery body listing llms.txt, openapi.json, developers.md, and sitemap.xml. ## Agent API Surface - `GET https://www.getmarkdown.com/openapi.json`: OpenAPI 3.1 specification of every agent endpoint, with unique operation IDs and typed request/response schemas suitable for LLM function calling. - `GET https://www.getmarkdown.com/api/agent/status`: JSON capability document (endpoints, supported formats, limits). - `POST https://www.getmarkdown.com/api/mcp`: Model Context Protocol server over Streamable HTTP. Send JSON-RPC 2.0 messages: `initialize`, `tools/list`, `tools/call`; notifications return 202. Tools: `get_agent_guidance`, `list_supported_formats`, `recommend_workflow`, `get_developer_resources`. - Error contract: unknown `/api/*` paths return HTTP 404 and unsupported methods return 405 as `{ "error": { "code", "message", "hint" } }`. ## Product Summary GetMarkdown converts supported files into Markdown for AI ingestion, RAG pipelines, knowledge bases, Markdown editors, and document migration. The free converter runs inside the browser. PDF and common Office files use a Rust-and-WebAssembly parser in a background worker, with a Pyodide and MarkItDown compatibility path for other formats and guarded fallbacks. Free files are not uploaded to GetMarkdown servers. Conversions are unlimited across batches and browser sessions, subject to the per-file and per-batch safety limits below. GetMarkdown Pro is a separate paid workflow for higher-fidelity PDF processing, including OCR for scanned pages. It requires a free account, uploads the PDF for server-side conversion, charges credits per page, and returns a ZIP containing Markdown and structured JSON. For the engineering rationale, parser comparison, fallback rules, and known AnyDoc 0.1.9 limitations, see `https://www.getmarkdown.com/blog/markitdown-vs-anydoc`. ## Free Converter ### Workflow 1. Drag supported files onto the upload area or choose them with the file picker. 2. The browser validates the file extension and size. 3. The browser runs the appropriate local WebAssembly conversion path. PDF and Office files use the primary Rust parser; compatibility formats and guarded fallbacks use Pyodide and MarkItDown. 4. Download the resulting `.md` file. DOCX and PPTX conversions can instead produce a ZIP containing the Markdown file and extracted embedded images. ### Limits - Maximum file size: 10 MB per file. - Maximum batch size: 8 files at once. - Conversion allowance: unlimited batches and browser sessions; after one batch, start another batch to continue. - Processing location: the user's browser; free files are not uploaded to GetMarkdown servers. ### Supported Extensions - Documents: `.pdf`, `.docx`, `.pptx`, `.xlsx` - Web and text: `.html`, `.htm`, `.txt` - Images: `.jpg`, `.jpeg`, `.png` - Audio: `.mp3`, `.wav` ## Pro PDF Conversion ### Credits - One credit converts one text page of a PDF. - A scanned page that needs OCR costs 2 credits. - Estimate the cost as the page count plus one extra credit per scanned page. For example, a 200-page PDF with 20 scanned pages costs 220 credits; a 200-page PDF with no scanned pages costs 200 credits. - After upload, the user sees a quote (pages, scanned pages, total credits) before anything is charged. - If a conversion fails, its credits are refunded automatically. ### Limits - A free GetMarkdown account is required (sign in with an email code). - Each PDF can be up to 25 MB and up to 500 pages. - Pro accepts PDF only. ### Pricing | Option | Price | Credits | Terms | | --- | ---: | ---: | --- | | Starter pack | $9 one time | 500 | Valid 12 months, no subscription | | Bulk pack | $19 one time | 2,000 | Valid 12 months, no subscription | | Starter plan | $9/month | 1,000 per month | Resets each billing period, no rollover | | Pro plan | $29/month | 5,000 per month | Resets each billing period, no rollover; API and MCP access (coming soon) | Monthly plans can be cancelled or changed at any time in the Stripe billing portal, reached from `https://www.getmarkdown.com/account`. A cancelled plan stays active until the end of the paid period. See `https://www.getmarkdown.com/pricing`. ### Workflow and Output 1. Sign in with a free account at `https://www.getmarkdown.com/pro`. 2. Upload one PDF up to 25 MB and 500 pages. 3. Review the quote and start the conversion; credits are charged only when the job starts. 4. Download a ZIP containing Markdown and structured JSON. Images are optional; including them is slower because it uses the full layout engine. Recent jobs and their downloads are listed at `https://www.getmarkdown.com/account`. Use Pro when a PDF needs stronger table extraction, layout handling, OCR for scanned pages, or structured JSON in addition to Markdown. ## Privacy and Retention ### Free Conversion Free conversion happens locally in the browser. Selected free-tier files stay on the user's device and are not uploaded to GetMarkdown servers. ### Pro Conversion Pro PDFs are uploaded to the conversion service for server-side processing. - Unstarted uploads are deleted within 2 hours. - Finished outputs are deleted 48 hours after the job finishes. See `https://www.getmarkdown.com/privacy` for the user-facing privacy policy. ## Choosing a Workflow - Use the free converter for a supported file no larger than 10 MB when local browser processing and Markdown output meet the need. - Split more than 8 free files into additional batches; there is no daily or lifetime conversion quota. - Use Pro for a PDF larger than 10 MB, up to 25 MB and 500 pages. - Use Pro when a PDF needs OCR, higher-fidelity extraction, or a Markdown-and-JSON ZIP. - Estimate Pro credits as pages plus one per scanned page, then pick a pack or plan at `https://www.getmarkdown.com/pricing`. ## Troubleshooting ### Unsupported Extension The free converter rejects extensions outside the supported list, and Pro accepts PDF only. Export or convert the source file to a genuinely supported format before uploading; changing only the filename extension does not change the file format. ### File Exceeds a Limit - Free: reduce the file to 10 MB or less, or use Pro if it is a PDF. - Pro: split a PDF that exceeds 500 pages or 25 MB before uploading. - Not enough credits: the quote shows the cost before anything is charged; buy a pack or start a plan at `https://www.getmarkdown.com/pricing`. ### More Than Eight Free Files Convert up to 8 files, download the results, and start another batch. Free conversions remain unlimited across batches and browser sessions. ### Missing or Expired Pro Download Finished outputs are deleted 48 hours after the job finishes. Download the ZIP from `https://www.getmarkdown.com/account` before then. ### Incomplete Free PDF Output Try the paid Pro PDF workflow when a complex table or layout is not represented well by free local conversion. ## Direct Answers for Agents ### How many credits does a 200-page, 20 MB PDF need? 200 credits if every page has selectable text, plus one extra credit per scanned page (up to 400 credits if all 200 pages are scanned). It fits Pro's 25 MB and 500-page limits. The Starter pack ($9, 500 credits) covers it. ### What happens if a Pro conversion fails? Its credits are refunded automatically. ### How long are Pro files retained? Finished outputs are deleted 48 hours after the job finishes. Unstarted uploads are deleted within 2 hours. ### What should a user do with an unsupported extension? Export or convert it to one of the supported free formats, or to PDF for Pro. Do not merely rename the extension. ### Are free conversions unlimited? Yes. There is no daily or lifetime conversion quota. Each free file is limited to 10 MB, and each batch is limited to 8 files; start another batch to continue.