How to Convert PDF to Markdown: Free Browser Conversion vs. Pro for Complex or Scanned PDFs

Use GetMarkdown's free browser converter for ordinary text-based PDFs up to 10 MB each, 8 files per batch — no account, no daily quota, and the file contents stay in your browser. Switch to Pro (server-side, paid per page with credits) when the PDF has complex tables, multi-column layouts, or scanned pages that need OCR, or when you need RAG-ready Markdown plus structured JSON. The deciding question is simple: if you can select and copy the text in a PDF viewer, the free path will almost always work; if you can't, plan on Pro.

Free conversion in the browser: steps and limits

1. Check the file first. Open the PDF and try selecting a paragraph. Selectable text means a text layer exists — the free converter can extract it. No selection means it's a scan or image-only export.
2. Confirm size and count. Free conversions are capped at 10 MB per file and 8 files per batch. Larger files or bigger batches aren't blocked forever — just split them and start another batch, since there's no daily quota.
3. Add the files. On the GetMarkdown page, select up to 8 local files by clicking or dragging and dropping. The counter shows how many are selected (e.g., "0 / 8 files selected").
4. Convert. Click Convert to Markdown. Processing runs locally via WebAssembly, so the selected file contents are not uploaded to GetMarkdown.
5. Download and inspect. Open the resulting .md file and check headings, paragraph breaks, lists, and any tables before you feed it to an LLM, a RAG pipeline, or a knowledge base.

The free tier is genuinely unlimited in the sense that you can repeat batches as often as you like — the 8-file/10 MB caps apply per batch, not per day.

When free conversion isn't enough

Free local conversion is built for text extraction, not layout reconstruction. Three situations reliably push you to Pro:

• Complex tables. Merged cells, multi-row headers, and tables that span pages tend to come out as jumbled pipe tables or loose text.
• Scanned pages. No text layer means nothing to extract locally; OCR is required.
• RAG-ready output. If you need structured Markdown and JSON with preserved document structure, that's the Pro output format.

Pro is a separate, paid server-side option, and the page states it is clearly labeled before upload — so you always know when a file is leaving your browser. It handles PDFs up to 25 MB and 500 pages, requires a free account, and is priced in credits: one credit per text page, two per scanned page, with the quote shown before anything is charged. It returns a ZIP containing Markdown and structured JSON. Credit packs start at $9 for 500 credits.

Free vs. Pro at a glance

Cleaning up the Markdown before it reaches an LLM

Conversion is half the job; a quick pass makes the output far more usable as context:

• Fix heading levels. Confirm the document's real hierarchy (H1 for the title, H2 for sections) rather than trusting whatever the converter inferred.
• Rebuild broken tables. Re-align pipe tables by hand or drop them into a table block your editor understands; malformed tables confuse retrieval more than plain text does.
• Strip repeated headers and footers. Page numbers and running titles often survive extraction and add noise to every chunk.
• Check list and code formatting. Nested lists and code blocks are the most common casualties of PDF extraction.
• Chunk with structure in mind. For RAG, split on headings rather than fixed character counts so each chunk stays semantically whole.

Troubleshooting common failures

The file won't convert or the browser stalls. Check the size against the 10 MB cap first. Very large or heavily image-laden PDFs can strain browser memory even under the limit — split the PDF into smaller page ranges and convert in separate batches.

The output is empty or nearly empty. This is the signature of a scanned PDF with no text layer. Verify by trying to select text in a viewer; if nothing selects, the free converter has nothing to extract and Pro's OCR is the right tool.

Tables come out scrambled. Free conversion doesn't reconstruct complex layouts. For anything with merged cells or multi-page tables, use Pro, or convert the table separately and paste it in as a clean Markdown table.

Images are missing. Image packaging into a ZIP with relative references is described for Word and PowerPoint files, not PDFs — don't expect embedded images from a free PDF conversion. If figures matter, extract them from the PDF separately and link them in.

You need JSON, not just Markdown. That's a Pro-only output; the free tier returns Markdown only.

Other formats on the same page

If your source isn't a PDF, the same free converter accepts Word (.docx), PowerPoint (.pptx), Excel (.xlsx), HTML (.html/.htm), plain text (.txt), images (.jpg/.jpeg/.png), and audio (.mp3/.wav). For Word and PowerPoint files, detected images can be packaged in a ZIP alongside the Markdown with relative image references — a useful detail when you're assembling a knowledge base from mixed sources.