Stories from the Japanese you know.
early access·feedback
How Mukashi creates stories.
Mukashi generates short Japanese stories using vocabulary you already know. The idea is simply to make reading practice easier to find at the right level.
What happens behind the scenes
Each story follows a structured pipeline. Mukashi starts with a selected vocabulary base, generates a story around it, reviews the result for fit and quality, and then presents the final output with reading tools that support comprehension without breaking flow.
Start with vocabulary you already know
Each story begins with a selected vocabulary base. Depending on your settings, that can come from WaniKani progress, JLPT vocabulary, or a blend of both. That vocabulary acts as the foundation for the story rather than a rigid checklist.
Write a story around your vocabulary
Mukashi generates a short Japanese story using the selected vocabulary as a guide. The goal is not to force every chosen word into the output, but to strongly bias the story toward language that should feel more familiar and readable.
Review for fit, difficulty, and naturalness
After generation, the story is reviewed to see whether it stays aligned with the intended level, uses vocabulary appropriately, and reads naturally as Japanese.
Read with tools that keep you in flow
The story is then shown in the reader, where you can toggle furigana, tap words to see meanings, or view an English translation if you want.
What Mukashi leans on
- Selected vocabulary as the foundation for each story
- Difficulty-aware generation
- AI review to catch mismatches and awkward output
- Reader tools that support comprehension
What Mukashi avoids
- Treating raw AI output as automatically good enough
- Forcing every selected word into every story
- Over-explaining internal heuristics
- Turning reading practice into a rigid drill