pyontan
EN
Blog Guides Pricing Download Discord
EN
Log in

Credits

Last updated 2026-10-11

pyontan stands on other people's work. Each dataset below stays under its own licence.

Dictionary

pyontan uses the JMdict, KANJIDIC2 and KRADFILE dictionary files. These files are the property of the Electronic Dictionary Research and Development Group, and are used in conformance with the Group's licence (Creative Commons Attribution-ShareAlike 4.0). They are read through the jamdict library.

Pitch accent

Pitch accent data comes from Kanjium (Creative Commons Attribution-ShareAlike 4.0), which is compiled from Wiktionary and other sources.

Example sentences

Example sentences and their translations come from Tatoeba and its contributors (Creative Commons Attribution 2.0 France), through the ready-made pairs at manythings.org/anki.

Word audio

Recordings of single words come from the yomitan-ultimate-audio collection and server by friedrich-de (animecards.site): NHK, Shinmeikai, Daijisen, JapanesePod101, Forvo and more, with speakers named where the recording has one. Words with no recording are read by Amazon Polly with their pitch accent.

Listening recordings

The recorded voices in Study's listening questions come from Mozilla Common Voice (Japanese), read by its volunteers and checked by other volunteers (CC0). The clips are trimmed and evened out in volume.

JLPT vocabulary lists

The N5 to N1 vocabulary lists are by Jonathan Waller (tanos.co.uk, Creative Commons Attribution), through open-anki-jlpt-decks by Jamie Sinclair (MIT). They are unofficial lists: the JLPT does not publish official ones.

Text analysis

Japanese text is split into words with SudachiPy and SudachiDict by Works Applications (Apache License 2.0).

Manga text detection

Whole-page manga scanning uses the comic text and bubble detector by ogkalu (Apache License 2.0), used unchanged.

Offline translation

The Offline translation engine runs LFM2-350M-ENJP-MT by Liquid AI (LFM Open License v1.0), through the ONNX export by onnx-community. It is downloaded from Hugging Face when someone installs it, and used unchanged.

Speech detection

Live Audio finds where speech starts and stops with Silero VAD by the Silero Team (MIT), used unchanged.

Icons

Most icons are from Lucide (ISC licence). The Discord and Patreon logos are from Simple Icons (CC0) and belong to their owners.

Roadmap Discord Chrome extension Contact us Terms Privacy Cookies Refunds Accessibility Copyright Credits
© pyontan 2026
We care about your privacy.

We use a small number of cookies to keep you signed in and remember your theme choice. We don't use any advertising or analytics cookies, and have no third-party ad partners. See our Cookie Policy and Privacy Policy for the full details.

Manage cookie options

pyontan doesn't run ads and has no advertising or analytics partners — here's what we use.

Strictly necessary Always on

Required for the site to work at all — keeping you signed in and remembering this cookie choice. Can't be switched off.

  • session — your signed-in session reference. Cleared on logout, or when the session ends.
  • cookie_consent — remembers the choice you make here, so we don't ask on every page. Expires after 6 months.

Remembers small choices you make, like light/dark theme and language, so you don't have to reset them every visit.

  • theme — your light/dark mode choice. Expires after 1 year if allowed; otherwise cleared when you close your browser.
  • lang — the language you picked for the site. Expires after 1 year if allowed; otherwise cleared when you close your browser.
Analytics & marketing None used

We don't use any analytics, advertising, or third-party tracking cookies on this site, so there's nothing here to turn on or off.