How to Import an Anki Deck Into Mochi
Mochi reads Anki packages, so the move is possible in one click - but a click is not always the right route. Here are the three ways across, what each one keeps, and the parts of an Anki deck that no route can carry.
Three routes from Anki to Mochi
Pick by how much structure your deck carries. A two-field vocabulary deck can go straight across; a five-field note type deserves the second route.
Route 1: import the .apkg straight into Mochi
- In Anki, right-click the deck, choose Export, and pick Anki Deck Package (.apkg). Leave scheduling information out - Mochi cannot read it anyway.
- In Mochi, open the deck sidebar, choose Import, and select the .apkg file.
- Check a few cards. Mochi maps the note's fields onto a card in the order they appear, which is what you want for Front/Back and rarely what you want for anything richer.
The trade-off: Zero tools, zero conversion. But you get whatever mapping Mochi decides on, and a note type with five fields tends to arrive as one wall of text.
Route 2: convert the .apkg to .mochi first
- Open the free Anki to Mochi converter and drop the .apkg in. The deck is read inside your browser - nothing is uploaded.
- The converter lists the deck's real field names (Word, Reading, Meaning, whatever your note type uses). Point one at Front and one at Back, and send the rest to Example, Notes, Tags or Ignore.
- Check the preview, then download the .mochi file and import that in Mochi instead.
The trade-off: One extra step, in exchange for control: you choose the two sides, and the extra fields arrive as readable Markdown under the answer instead of being flattened or dropped.
Route 3: regenerate the deck for Mochi
- Open the Mochi generator, set your language pair, and describe the same material the old deck covered.
- Generate, review the cards, and download the native .mochi package.
- Keep the Anki deck around until you are sure the new one covers everything you still need.
The trade-off: Costs a credit and abandons your old cards, but you end up with example sentences and Markdown-ready formatting instead of a decade-old CSV import.
What survives the move
Anki and Mochi are not the same shape. Anki has named fields, note types and a media archive; Mochi has two Markdown sides. Here is what that difference costs, before you find out one card at a time.
| What | Verdict | Detail |
|---|---|---|
| Card text | Moves | Front and back arrive intact. HTML from Anki is stripped down to plain text and line breaks, so bold-by-markup does not survive as bold. |
| Extra fields | Needs a decision | Mochi cards have two sides, not named fields. Example, Notes or Reading either get folded into the back or dropped - the converter folds them in as Markdown so nothing disappears silently. |
| Tags | Moves, differently | Anki tags become hashtags at the end of the card in Mochi, which is how Mochi picks them up. Anki's :: hierarchy flattens. |
| Audio and images | Stays behind | Media lives inside the .apkg archive and is not carried across by the converter. It tells you how many files the deck contained so the loss is never a surprise. |
| Review history and scheduling | Stays behind | Neither route moves your intervals or ease factors. Mochi schedules every imported card as new, and there is no format that would bridge the two algorithms honestly. |
| Cloze deletions | Needs rewriting | Anki writes {{c1::like this}} and Mochi does not read it. Cloze cards arrive as literal text and have to be redone in Mochi's own syntax. |
| Card templates and styling | Stays behind | Anki's note types and CSS have no equivalent to import into. Mochi renders Markdown, and that is the whole formatting model. |
Still deciding whether to move at all? Mochi vs Anki compares the two apps on scheduling, formatting and daily use. If you are starting fresh rather than migrating, the Mochi generator writes cards in the format Mochi wants from the start, and CSV to Mochi handles a spreadsheet source.
Rebuilding fields, cloze and structure in Mochi
The import is the easy part. The work that decides whether the migrated deck is worth reviewing is what you do with the pieces Mochi models differently.
A serious Japanese or Chinese deck has Word, Reading, Meaning, Example and Notes. Mochi gives you a front and a back, so the design question is which field earns the front and what order the rest go in underneath. The usual answer: the word alone on the front - recognition is the skill you are training - then reading, meaning, example and notes on the back, separated by blank lines so Markdown renders them as distinct paragraphs. Putting the reading on the front turns a recall card into a reading-aloud card, which is a different exercise entirely.
Anki hides text with {{c1::...}} inside a Cloze note type; Mochi hides it with its own bracket syntax on a normal card, and supports several deletions in one card the same way. There is no automated bridge, so the practical approach is to search your Anki collection for note:Cloze before exporting, decide whether those cards are worth rebuilding at all, and rewrite the survivors by hand in Mochi. A deck that is mostly cloze is usually a deck worth regenerating rather than migrating.
Anki leans on a deck tree plus tags; Mochi leans on decks plus tags with far less hierarchy. An Anki deck named Spanish::Verbs::Irregular arrives as a flat name. Before importing, decide whether that structure was carrying real meaning - if it was, split the export into separate decks per branch and import them one at a time, which is cheaper than reorganizing hundreds of cards afterwards.
If your deck's value is in its audio - a sentence bank with native recordings - moving it to Mochi means losing the thing that made it good, and the honest advice is to keep studying it in Anki. If the media is decorative images on a vocabulary deck, drop them: the card still works. There is no in-between where the text migrates and the audio follows later, because the media filenames in the text refer to files Mochi never received.
Plenty of people keep Anki for the deck with 8,000 mature cards and use Mochi for everything new. That works as long as one app owns each subject: two copies of the same vocabulary on two schedules means reviewing everything twice and trusting neither. Draw the line by material, not by feature - old deck stays in Anki forever, new material starts in Mochi - and generate new decks directly in the .mochi format so nothing needs converting at all.
Premium access includes:
Frequently asked questions
Yes. Mochi reads Anki packages, which is why route 1 exists and needs no tools. Converting first is about control over the mapping, not about compatibility.
No, on either route. Every imported card starts as new in Mochi. If the scheduling history is what you value most, that is an argument for keeping the deck in Anki.
No. The converter moves text only and reports how many media files the deck held. A deck whose point is its audio is better left in Anki.
You choose. The converter shows the deck's real field names and lets you pick the front and the back, sending the rest to the card back as Markdown or dropping them.
Often, yes - for small or old decks. Regenerating costs a credit and gives you cards written with example sentences and Markdown in mind. For a large curated deck you have been growing for years, convert instead.
Not with these tools. Mochi exports Markdown, which you can reshape into a spreadsheet and then run through the CSV to Anki converter, but there is no one-click route.
Keep reading
Related guides that build on what you just read.