Anki deck to CSV,
Quizlet or PDF

What you download

Pictures and sounds stay in Anki: the page counts them and tells you how many, and each row keeps the text around them. Review history and scheduling stay in Anki too.

Turn it into a Quizlet set

1. Open the deck above. Under Choose the fields, mark the field that should be the term (the front of the card) and tick the ones that belong on the back.
2. Choose Copy for Quizlet. The text is now on your clipboard.
3. On Quizlet, create a new set and choose Import. Paste the text into the box.
4. Keep Tab between term and definition and New line between cards, check the preview, and create the set.

Quizlet’s Import box reads a tab, a comma or a dash between term and definition, and a new line or a semicolon between cards (Quizlet Help Center). This page writes a tab and a new line because a comma or a semicolon can appear inside a meaning. A deck with clozes reads better in Quizlet with Cloze notes: show each gap as ___ ticked: the sentence with its gap becomes the term, and the answer comes first on the back.

Open it in Excel or Google Sheets

• Excel: open the downloaded .csv. If every row lands in one column, your Excel expects semicolons (common where a comma is the decimal sign, as in German, French or Spanish settings): download again with Semicolon, or use Data, From Text/CSV and choose the separator there.
• Google Sheets: File, Import, Upload, and choose the file. Sheets finds the separator itself.
• Numbers and LibreOffice: open the file; both read UTF-8 CSV and TSV.
• Each note is one row. The first column is the term you chose; note types that share a field name, such as Basic and Basic (and reversed card), share that column.

Going the other way, from a spreadsheet back into Anki? Save it as CSV and use the free Quizlet to Anki converter, which writes a file Anki imports.

How this page reads your deck

• An .apkg (one deck) or .colpkg (a whole collection) is a zip file holding Anki’s SQLite database, plus your pictures and sounds. The page opens the zip in your browser and reads only the database.
• Anki 2.1.50 and later can save that database in a smaller, compressed form (collection.anki21b, zstd); current Anki does so unless “Support older Anki versions” is ticked when you export (Anki 2.1.50 changes, Anki manual, exporting). The page reads it, the older collection.anki21 and the Anki 2.0 collection.anki2.
• From the database it takes the notes, the names of their fields and their decks. A note sits in the deck of its first card; a card moved to a filtered deck counts in its home deck. Choosing a deck includes its subdecks.
• Fields are made plain text: HTML tags go, a line break stays a line break, & and other entities become their characters, [sound:…] and pictures go, and a reading in a ruby (furigana) is left out. A cloze shows its answer, or a gap if you ask.
• Which field starts ticked follows Lexpad’s importer: fields named for audio, pictures or video start unticked, and so does any field that is empty in every note. You can change every choice.
• The readers are sql.js 1.14.2 and fzstd 0.1.1 (both MIT licence), served from lexpad.app and fetched only after you choose a file. The zip is unpacked by your browser itself.
• Nothing is sent or saved: no upload, no cookies, no storage in the browser. Close the tab and the page forgets the deck.

Tested on 25 October 2026 with decks we made in Anki’s own code (version 26.9.3): .apkg and .colpkg in the current and the older formats, an Anki 2.0 package, cloze notes, a note type with its own fields, subdecks, tags, pictures and sounds, Arabic and Japanese. Each deck’s rows were checked against Anki’s own plain-text export of the same deck, and the page was run in the engines of Chrome, Safari and Firefox, watching that no request carried the deck.