macOSWindowsiOSAndroidHub
Scanned PDFs and OCR
Improve Mind chat on image-only or low-text PDFs with server-side OCR.
Before you start
- PDF source uploaded to a Mind notebook.
Steps
- 1
Check the source Quality score in Sources — low scores on PDFs often mean little extractable text.
- 2
FlyType runs OCR automatically on scanned PDFs when Vision is configured (see Phase 76 ingest).
- 3
Large scans may stay in Processing for several minutes — wait for Ready before chatting.
- 4
If ingest failed but body text exists, open the source preview and tap Re-process.
- 5
Re-upload a higher-resolution scan if OCR text is still empty or garbled.
Common mistakes
- Prefer searchable PDFs when you control the export (Print to PDF with text layer).
- Trial plans have shorter media caps; Pro allows longer OCR jobs.
- Use folder import preview to skip duplicate scans already in the notebook.