Adding pages

You add new pages by uploading scanned images or by dropping in a PDF. The book takes it from there — it straightens scans, reads any text on the page, finds faces, and writes a first-draft caption. Everything is editable afterward.

Starting an upload

Open the book and press the Add Pages button in the top bar, next to Home and Chat. The upload window appears. (You can also reach it from Admin › Add Pages.)

In Simple View, there’s an Add Pages button right in the toolbar instead.

Choosing files

1. Drag and drop files into the dashed box, or click the box to browse.
2. JPEG, PNG, WebP, AVIF, TIFF, HEIC, and PDF files are all supported, and you can mix them in one batch. (HEIC is the format iPhones save photos in by default.)
3. Each selected file shows up as a small preview below the dropzone. PDFs show as a single “PDF” tile (expanded into one book page per PDF page on upload). HEIC and TIFF files show as a labeled tile because browsers can’t render those types directly — they’re converted to a standard image on the server.
4. Have a whole folder of scans? Use the “Choose a folder” link under the dropzone, or drag the folder itself onto the dropzone. Every supported file inside the folder (and its subfolders) gets added; anything else is skipped automatically.
5. Individual photos must be under 30 MB each (documents under 25 MB; PDFs can be much larger — up to 250 MB). If a file is too large, it’s flagged in the preview with a note about how to shrink it, and the Upload button stays disabled until you remove or replace it. A very high-resolution scan or TIFF can exceed the photo limit — re-saving it as a JPEG usually drops it to a few MB with no visible loss. (Large PDFs, audio, and video upload by a different route, so their limits are much higher.)

Sort order

The pages will be inserted into the book in the order you pick:

• Default — the order in which you dropped or selected them.
• Name — alphabetical by filename. Useful if you named scans like page_01.jpg, page_02.jpg.
• Modified Date — oldest file first. Useful for batches scanned in chronological order.

Where the new pages land

Leave the Insert at position field blank to append to the end of the book. Or type a page number — the new pages slide in starting at that position and everything after shifts down.

Tell the book what you know

The “What do you know about these pages?” box lets you add context the AI can’t guess from the image alone — an address, a building, an event, a year. Whatever you type here applies to the whole batch you’re uploading and is used when the book writes its first-draft captions.

It’s most useful for photos with no writing on them. Without context, the AI might write “an old building, likely from the early 1900s.” Tell it “this is St. Mary’s Church, where Grandpa’s funeral was held,” and the caption can say so.

About PDFs

Drop a PDF in just like an image. When you press Upload, each page of the PDF becomes a new book page. If the PDF contains typed text (e.g., a born-digital document, or an OCR’d scan), the book preserves it:

• For pages that are pure typed text (no scanned image inside the PDF), the typed text becomes the page’s transcript directly.
• For pages that contain a scanned image (a photo, handwriting, etc.), the typed text is treated as supplementary — it stays attached to the page and the AI sees it while reading the page, but the visible image is still the canonical content.

Long PDFs are rendered one page at a time in the background, so the book stays responsive even with hundreds of pages.

Typed documents (Word and text files)

You can also add .docx and .txt files — a typed family history, a transcribed memoir, a letter saved from a word processor. These become text pages: there’s no scanned image, the typed words themselves are the page. Long documents are split into pages of roughly a typed page each, and the original file is kept safely in storage exactly as you uploaded it.

Text pages show a typed-page card in the book view, work with search and chat like any other page, and can be grouped into a reading sequence so the whole document reads continuously.

What happens during processing

Once you press Upload, the form is replaced with a progress panel. For each page, the book runs a series of steps:

• Upload — the image is saved.
• Auto-crop & rotate — crooked scans are straightened.
• Thumbnail — a small preview is generated.
• Text extraction — any writing on the page is read into the transcript.
• Face detection — faces are found and added to People & Faces.
• Caption — the AI writes a first draft caption.

Each step shows a check as it finishes. You can close the dialog — the work continues in the background, and the pages will appear in the book as they’re ready.

If an upload fails

If one step fails (a scan is too small, the file is corrupt), that page pauses with an error. The rest of the batch keeps going. You can retry the failing step from the admin panel, or delete the page and re-upload.