Keep your files with your pages, and find the words inside them.

Read them where they are

• PDFs in a full viewer: pages drawn as you scroll, selectable text, find, zoom and full screen.
• Documents: Word (.docx, .odt), slides (.pptx, .odp), spreadsheets (CSV, TSV, Excel, OpenDocument) with sorting, “Use in a code cell” and “Make a chart”, and EPUB books that remember your place.
• Data and code: JSON, YAML, TOML and XML as trees; source files highlighted; Markdown rendered.
• Photos in a lightbox; audio with a waveform and chapters; video with subtitles, picture-in-picture and speed. A timestamp like [03:21] under a recording jumps to that moment.
• SVG and HTML files open in a sandbox; their scripts run only if you ask.

Search inside them

Text in ordinary PDFs, Word documents and slides is searchable straight away. Scans and photos become searchable once you turn on text recognition in Settings → Documents: it downloads about 15 MB once, from Vellum, and reads the images on your device.

Results point at the page of the file, inside the page that holds it, and the assistant can cite those pages.

Private by default

File bytes live in the browser’s private file system on your device, next to the pages. Images on the web are never loaded by a page; Vellum offers to download them into the page when you ask.

In the app

• Files and photos Keep PDFs, photos and documents with your pages.
• All your files Every file in one list, with the page it lives in.
• Text in photos and scans Make words in images searchable. Off until you want it.
• Search everything Find any word, even inside your files.
• Pictures from the web Save web images into your pages, only when you ask.

Questions