Turn scanned PDFs into clean, searchable books.
Import your scanned PDF, run OCR, then read it as real text β with page breaks and a layout that stays close to the original.
Key features
Open from Files, iCloud Drive, or other apps. Keep everything organized in a simple library.
Designed for long, multi-page scans. Keeps text aligned with each page for easy navigation.
Read as clean text with page breaks preserved β closer to an ebook, without losing context.
Jump to keywords instantly. Copy and share passages, or export the full OCR text.
Listen to long documents with system text-to-speech when enabled.
Your files stay on device unless you choose to share or export. No ads, no tracking.
Quick start
Tap Import and select a PDF from Files, iCloud Drive, or Share Sheet.
Choose full book OCR or only selected pages. Processing runs as efficiently as possible.
Tap TXT to read reconstructed text while keeping page breaks familiar.
Search for keywords, copy passages, and export text for notes or other tools.
Local AI with Ollama (optional)
Use Ollama to summarize, translate, or ask questions β while keeping content on your own machine. If you connect over plain HTTP, traffic may not be encrypted. Use a trusted network or HTTPS where possible.
Smart summarization workflow
Match the plan to your Ollama server with a 4Kβ128K context length slider, so requests fit what your hardware can actually handle.
OCRBook previews the exact strategy and number of server calls a summary will take β no surprises once it starts.
When a document is too long for one response, OCRBook splits it into parts, summarizes each one, and merges the results automatically.
Ask for a quick recap or a much longer, detailed summary β even for a very long book β and the workflow adjusts itself to match.
Attach the whole document to chat. If it's too long for the context window, OCRBook scans it part by part for the answer instead of cutting it off.
Summarization and full-document chat run as a visible, step-by-step workflow that you can cancel at any point.
The document fits your chosen context window, so one request produces the full summary.
The document is split into parts, each part is summarized, then the partial summaries are merged into one.
Partial summaries are merged over several rounds before the final summary is written.
For long, detailed summary requests, the result is written as multiple sections and stitched together on your device.