You have two decks that cover the same exam. Before you merge them, delete one, or decide what to study, you want to know one thing: how much do they actually overlap? Anki cannot answer this. There is no diff view, no overlap report, no built-in comparison of any kind.
The manual ways, ranked by pain
Sort and eyeball. Open both decks in the browser, sort by the sort field, and scroll through them side by side looking for matching facts. This works for about fifty cards. Nobody has ever done it for five thousand.
Export and diff. Export both decks as plain text, open them in a spreadsheet, and compare. You immediately hit the formatting problem: HTML, cloze markers, and line breaks make text comparison noisy, and reworded cards still do not match. You spend the afternoon cleaning data instead of studying.
Import and Find Duplicates. Import deck B into a staging deck in the same profile, then run Edit > Find Duplicates. This catches exact first-field matches and nothing else. Two cards testing the same fact with different phrasing sail through. It also pollutes your collection with an import you have to delete afterward.
All three methods share the same flaw: they compare text, not meaning. Two cards can test the identical fact with zero words in common, and every manual method misses them.
Why overlap matters
Overlap is not trivia. It answers real study questions. If 80 percent of a shared deck overlaps with cards you already made, you do not need the shared deck — you need the 20 percent that is new. If you are deciding whether a second resource is worth its price, the overlap number is the answer. If two study partners want to split card-making without duplicating effort, the overlap check keeps them honest.
The better way: overlap as an operation
Upload both .apkg files and pick Overlap. The comparison runs semantically, so cards that test the same fact match even when the wording differs. The result is a deck containing only the facts covered by both decks — the shared core, downloadable as a single .apkg.
Use it as a diagnostic before the real decision. High overlap with your own deck means the shared deck adds little. Low overlap means it is worth importing. Either way you know in minutes instead of after an evening of spreadsheet work.