PDF Reader & OCR
The PDF Reader is where every scanned document, downloaded form, and Talking Camera capture eventually ends up, and it's built to handle both the easy case — a PDF with clean, selectable text — and the hard case, a scanned image with no text at all. Getting started is simple: "Upload PDF" opens a file picker, or "Open Document Scanner" launches the camera to capture a fresh page on the spot. Everything you open is kept in a library beneath these buttons, with a "Clear Library" option if you'd like to tidy up.
Once a document is open, moving through it is entirely gesture and voice friendly: swipe to move between pages, or use the dedicated "Go to next page" and "Go to previous page" accessibility actions your screen reader exposes directly. The page indicator, reading something like "Page X of Y," doubles as a button — tap it to jump straight to any page number rather than paging through one at a time. A "Search in PDF" option in the header lets you locate a specific word or phrase across the whole document instantly.
A bottom bar keeps your core tools within constant reach: the shared hands-free playback controls for listening to the current page read aloud, a "Bookmark" toggle (which becomes "Bookmarked" once set) for marking your place, and an "OCR" button for scanned pages that don't yet contain readable text. Tapping OCR opens a choice of three strategies: "Quick OCR" for a fast pass, "Foundational OCR" — recommended specifically for image-heavy or lower-quality scans — and "AI OCR," the most accurate but slowest option, powered by cloud AI. When running AI OCR, you can choose its scope: just the Current Page, a Full Scan of up to fifty pages at once, or a Custom Page Range you define yourself. If a PDF is password-protected, the reader simply asks for the password before opening it.
The standout feature here is "Ask AI," which opens a dedicated "Ask about this document" panel offering a few quick-prompt suggestions alongside a free-text field for your own question. Ask something like "What are the key dates mentioned in this contract?" and the question, together with the document's content, is handed off into Tech Assistant AI, where you get a genuine, conversational answer rather than having to read the whole file yourself to find it.
A "More Actions" menu covers everything else you might need: copying, analyzing, or translating text with a choice of scope (current page, the entire file, or a custom range), sharing the extracted text or the document itself, sharing the current page as an image, exporting content, sending the current page or full file straight to Tech Assistant AI, viewing all your saved bookmarks in one list (tap any entry to jump straight there), checking file information, renaming the document, removing it from your library, or deleting a previous OCR result if you'd like to try again. If OCR is set to run in the background rather than on the current page, a tappable progress banner keeps you updated — "Processing in background: Page X of Y" — so you can keep reading while the rest of the document catches up.