Making a bilingual PDF you can actually study from

Two different jobs

People say "bilingual PDF" for two things that need opposite solutions.

Reading a text — an article, a chapter, a story. Here you want the original and translation aligned so your eye moves easily between them, and you do not care that the layout no longer matches the source.

Working with a document — a paper, a form, anything with columns, tables or figures. Here the layout is part of the meaning, and a tool that reflows it into one column has destroyed the thing you were trying to read.

For reading: aligned columns

Translation runs in batches rather than all at once, so a long text becomes readable from the beginning before it has finished. The reader also shows coverage — what percentage of the text is words you already know — which is more useful than a reading level, because it is about you rather than an average reader.

1. Open the PDF in Chrome.
2. Press "Make a bilingual book" in the bottom-right corner.
3. The document opens in two columns, with paragraphs aligned so the original and its translation begin at the same height.
4. Hover a sentence to highlight its counterpart in the other column.
5. Click any word for the meaning it has in that sentence, and save it if it is worth keeping.

For documents: keep the original layout

The second button takes the opposite approach. It draws each page exactly as the file specifies, then lays a transparent, precisely aligned text layer over the top.

What you see is the untouched document, with columns, tables, figures and typography intact. What you interact with is real text: every word has its own box, so it can be clicked, hovered, and coloured by how well you know it.

This is the mode for academic papers and anything where reflowing would destroy the point.

Limits worth knowing before you start

Fifty pages in the reader

Longer documents need splitting first.

No OCR

A scanned PDF with no text layer cannot be read by either mode. If you cannot select text in Chrome’s own viewer, neither tool can help.

Local files need one extra permission

For a file:// URL you have to allow access at chrome://extensions, or the buttons never appear at all.

Permission is requested at the button

Document access is asked for the first time you press one of these, never at install. Until then the extension has no access to your documents.

Importing something that is not a PDF

Open the reader without a selection and you get an import field. It accepts .txt, .srt, .vtt, .md and plain pasted text, up to 400,000 characters.

Subtitle formats are worth calling out: the timings and sequence numbers are stripped, so you get readable prose rather than a list of timestamps. That turns a subtitle file you downloaded elsewhere into something you can actually read through.

What to do with the words

The point of reading bilingually is not the reading, it is what you keep. A word clicked in a PDF saves exactly as one clicked in a subtitle does — into the same deck, with the sentence it came from attached.

Both modes, their limits and the 50-page cap are described in full on the bilingual PDF feature page.

The card records the document name rather than a video title, so it never claims to come from a video it did not.