Vocabulary Builder keeps every word you looked up on the Kindle and gives you no export button. The words sit in a SQLite file called vocab.db on the device itself. Here is how to pull them out with their sentences, and what to do about everything you read that is not on a Kindle.
Plug the Kindle into your Mac with a USB cable and it mounts as a drive. The file is at system/vocab.db, and system is hidden, so press ⇧⌘. in Finder to reveal it. Copy the file somewhere else before you touch it.
It is an ordinary SQLite database. WORDS holds the words, LOOKUPS holds the sentence each one was found in, and BOOK_INFO holds the titles.
macOS ships with sqlite3, so nothing needs installing. Point it at your copy of the file:
sqlite3 -csv vocab.db \
"SELECT w.stem, l.usage FROM WORDS w
JOIN LOOKUPS l ON l.word_key = w.id;" > kindle-words.csvThat gives you a CSV with the word in the first column and the sentence it came from in the second, which is exactly the shape Anki wants on import.
Swap w.stem for w.word if you would rather keep the inflected form you actually met.
Amazon has never shipped an export, so the file is the only route. Worth knowing before you rely on it: the list only records lookups made on the Kindle itself.
Everything you read somewhere else is invisible to it. Papers, PDFs, websites, subtitles, an ebook open in a desktop reader. For most people that is the majority of their reading.
The CSV above is already the shape Anki wants: word in the first field, sentence in the second. In Anki, File, Import, pick the file, set the separator to comma, and map field 1 to Front and field 2 to Back. Or map the sentence to Front and the word to Back if you would rather be tested on recognising it in context, which is the harder and more useful direction.
If a sentence contains a comma the quoting handles it, because -csv quotes the fields for you. That is the whole reason to use sqlite3 rather than opening the file in a viewer and copying columns by hand.
It only exists on the e-readers. The Kindle apps for iOS, Android, Mac and Windows have never had it, so if you read in an app there is no list to export and nothing on the device to find.
On a Paperwhite or Oasis it is a tile in your library rather than a menu item, and recent firmware moved the way in: open a book, tap the right hand sidebar, then Flashcards. It also has to be switched on, under the three dot menu, Settings, Reading Options, Vocabulary Builder. One more catch that costs people an afternoon: the vocabulary database syncs through your Amazon account, so an unregistered Kindle will not build a list at all.
Check the same three things first: Vocabulary Builder switched on, the Kindle registered, and a dictionary actually downloaded for the language you are reading. A missing dictionary is the quiet one, because the lookup popup still appears and simply has nothing to record.
There is also a limit nobody advertises. Readers on Amazon's own forum report older entries being dropped once the list gets somewhere around two thousand words. Treat that as a reason to export sooner rather than later: the words at risk are the oldest ones, which are the ones you have had longest and are least likely to remember collecting.
LeafyApp is the same idea pointed at a Mac screen. Press ⌥A, box the word, and it saves the word together with the sentence it appeared in, which is the pairing that made the Kindle list useful in the first place.
From there it exports to Anki, CSV, Excel, JSON or PDF, with the card layout previewed before the file is written. Same data as the Kindle list, over more of your reading.
LeafyApp does not import that file today. The sqlite line above is the way to do it, and it takes about a minute. If enough people want the import built in, say so on Discord and it goes on the list.
Your library stays on your Mac. No account, no sync server, no ads. The privacy page spells out what is sent and what is kept.
Free while LeafyApp is in beta. One download, macOS 14 and up.
Also worth a look: building an Anki deck out of what you read, looking up a word on Mac, or sentence mining on a Mac.