Anki Listen Mode

An Anki add-on that adds a Listen (Hands-free) mode: it reads through a deck (or just the due queue) out loud, card by card, advancing on its own — no clicking, no grading, just listening. If a note type happens to have an example-sentence or synonyms field, those get read too.

Why this exists, and how it fits into Anki

Anki's core has no concept of "words" — it's a general-purpose spaced repetition tool, and decks are entirely user-defined. So this add-on is built to be generic:

• Hands-free playback works on any note type. It reads the rendered question, then the rendered answer (with the repeated front text stripped out, since most templates restate {{FrontSide}}).
• Example sentence / synonyms are opt-in: if a note type happens to have a field named Example (or Sentence, Usage, ...) and/or Synonyms (or Synonym, ...), those get shown and read after the answer. Field names are configurable in case yours differ (config.md).
• Speech is not reimplemented here — it hooks into Anki's own {{tts xx_YY:Field}} engine (anki.sound.TTSTag / aqt.sound.av_player), the same mechanism card templates already use. No API keys, no network calls, no bundled TTS engine.
• Sync is untouched. This add-on only reads cards and notes (card.question(), card.answer(), note.items()) — it never writes to the collection, so it has zero interaction with Anki's sync system. It works identically whether or not you have sync configured.

Install

1. Copy this folder into your Anki add-ons directory as listen_mode: - Windows: %APPDATA%\Anki2\addons21\listen_mode - macOS: ~/Library/Application Support/Anki2/addons21/listen_mode - Linux: ~/.local/share/Anki2/addons21/listen_mode
2. Restart Anki.

(Packaging as a .ankiaddon for AnkiWeb is just zipping this folder's contents at the top level — manifest.json at the zip root.)

Use

• Tools → Listen (Hands-free)… — pick a deck, whether to include the whole deck or just what's due, and a play order, then it opens the player: Play/Pause, Prev/Next, and it keeps going by itself.
• Browser → select some notes → Notes → Listen to Selected Notes… — same player, seeded from an arbitrary search/selection instead of a deck.

Neither entry point touches scheduling or grades — it's closer to Anki's own Preview than to Study, just hands-free and read aloud.

Configuration

See config.md — speaking rate, gap between fields/cards, TTS language, and the field-name lists used to detect example/synonyms fields.

Development

textutils.py (HTML→text cleanup, front/back de-duplication, field matching) is plain Python with no anki/aqt import, so it's unit tested without needing an Anki install:

python -m unittest discover -s tests -v

The rest (listen.py, __init__.py) is real Qt/aqt code and was checked against Anki 26.8.1's actual anki/aqt source (not just docs) for every API it calls, and exercised both by loading it in a real, isolated Anki profile (Anki.exe -b <empty base dir>, so it never touches a real collection or triggers a sync) and by running its exact note/card-handling logic against a real anki.collection.Collection with a custom note type. It hasn't been driven through the Qt UI itself (clicking through the dialog), since that needs a human at a screen — give it a try and open an issue if something in the dialog itself misbehaves.

Contributing

This is a standalone add-on, not a patch to Anki's own core (ankitects/anki is a large Rust + Python + TypeScript/Svelte monorepo with its own build toolchain — a good add-on is the normal way to ship a new feature like this without going through that). Contributions, issues and PRs against this repo are welcome. If a maintainer ever wanted a piece of this folded into Anki core directly, that would be a separate, from-scratch implementation in Rust/Svelte against their codebase and conventions.

License

AGPL-3.0, matching anki/aqt. See LICENSE.