Known Issues

Things we know don't work well yet, and what you can do about them. The page below is a curated summary of the issues most players run into.

See every reported issue, live Browse and follow all reports — including ones raised by other players — on our public feedback tracker on GitHub. New reports and status changes show up there in real time.

Status legend

Investigating We're looking into it. A fix is in progress.
Mitigated A workaround exists. Root fix is planned or not practical.
By design This is expected behavior or out of our control.

Japanese audio playback can mispronounce kanji

Investigating

What you'll see When you press the audio button on a Japanese sentence, the text-to-speech engine reads kanji with incorrect or inconsistent readings. In long sentences, this can result in playback that does not sound like natural Japanese.

Why Japanese kanji-to-reading conversion is hard — the same character can have multiple valid readings depending on context, and the current TTS pipeline does not yet use a full Japanese morphological analyzer for pronunciation.

Workaround For now, audio playback is most reliable on single words saved to your dictionary rather than full sentences. Pronunciation for learning practice is best verified against a native dictionary entry.