About the project

A trainer for spoken Japanese. You go through a dialogue line by line: your partner's lines are played back to you, and the lines of your own role have to be said out loud. Recognition runs in the browser, and until a line comes out clearly the player will not move on. The point is forced output.

Who it is for

People studying Japanese on their own at N5–N3 who have run into the gap between "I understand" and "I can say it".

What it takes to work

A browser with Web Speech API support and a microphone. Nothing to install, no extensions.

Availability

The project runs entirely on free tiers and is offered as is: availability is not guaranteed. The database goes to sleep without traffic, the daily allowance of emails is limited, and the hosting has cold starts. This is not a service with an uptime commitment.

Feedback

Found a bug, an inaccurate translation, or just want to say something — write to the developer.