PDF / MARKDOWN / KNOWLEDGE BASE
How to import PDF documents into Notion, Obsidian and Logseq — a PDF to Markdown guide
Turn a PDF into an editable note, review the result, and bring it into your knowledge base. One conversion button starts the process; importing into your note app is a separate step.
1. Check your PDF
Try selecting text in the original PDF. If it works, use text extraction. For a scanned PDF, enable OCR and select English, Chinese or French to match the document. Files can be up to 25 MB and 200 pages.
2. Convert locally
Open PDF to Markdown, drop in your file, and choose heading detection or page separators as needed. Keep OCR enabled for pages without selectable text. Click Convert to Markdown. Choose the language manually and Fast recognition for free OCR. Automatic and high-accuracy modes are planned paid features, not yet available. Free tools need no login: this browser has 10 successful conversions and 30 OCR pages per day, shared across tools.
3. Review before downloading
Use Source and Preview to check headings, line breaks, names, numbers and reading order. Edit the source directly, then download .md. Images and table layouts are not preserved; complex columns and handwriting need careful review. Keep your original PDF. Leave optional YAML metadata off unless your workflow expects it.
4. Import into Notion
On desktop or web, go to Settings → Import → Text & Markdown and select your .md file. Review the new page and move it to the appropriate workspace. Importing here sends the document to Notion.
5. Add to Obsidian
Copy the .md file into a folder inside your existing vault, then open the note in Obsidian. Back up your vault and avoid overwriting a note with the same filename.
6. Add to Logseq
For a file-based Markdown graph, place the .md file in its pages folder and reopen the graph. Review the block outline. Database graphs have a different import workflow; follow the documentation for your version.
Troubleshooting and useful next steps
For empty output, check whether the page is a scan and whether OCR is enabled. For inaccurate recognition, use a clearer scan and the correct language. When the daily OCR allowance is exhausted, try again the next day. For an image or screenshot, use Image OCR instead. RAG JSON exports are for custom ingestion pipelines, not direct note imports; they do not contain embeddings.
Official documentation: Notion · Obsidian · Logseq (Markdown)