Lingon

I built this because I want to learn Swedish and there are too many good books in the world that I am yet to read. So I decided I wanted to bring those books down to my level instead of spending time on boring study materials 😇

Features#

You can import any EPUB, or paste plain text or Markdown into Lingon. It doesn’t have to be a book, either. You can tag the texts you import as an article, lyrics, a nursery rhyme, or anything your heart desires.

Translate and adapt to any language and level#

You can translate directly from the original, or select your level and Lingon will adapt the text so you can read it without too much difficulty. Dracula in Swedish at A1? We got you!

If the book is already in your target language, you can always try to read the original. Too hard? Try to bring it down. Was it too easy? Try a level up!

AI generated adaptations are done per chapter. This does a couple of things. It makes sure we are not burning tokens and translating a bunch of stuff you might not even read. And it also allows your level to fluctuate as you read through the book.

Narration, recording and shadowing#

You can listen to the book in your target language, record yourself while you read, or put on some headphones to shadow (listen to the book and repeat what you hear).

The recordings are saved as plain files in your file system. You can listen to them in Lingon or in any other program of your choice. You can also send them to a friend to critique your pronunciation and even go through the embarrassment of listening to yourself 😊

Save for later#

Words, passages, notes, and AI chats are all available in the sidebar. But you can also view them as collections on a page containing everything you have saved across different chapters, books, and languages.

Word definitions#

Click a word to save it. My idea here was to keep it very low friction. As shadowing is my modus operandi, I don’t want to worry too much about words I don’t understand. If I get the gist of the passage I’m reading, I can save any unknown words to review later without interrupting my flow.

Passages#

Passages can be highlighted and saved with a note. Both saved passages and words are linked to the exact position in the book, so you can easily go back to look at them in context.

Notes#

There is also a side panel where you can write notes as you read. They are linked to the book chapter and saved as plain Markdown files in the file system, so your notes are portable and can be accessed from anywhere.

AI threads#

When in doubt, you can bring in a little bot to help you clarify whatever is confusing you at the moment.

We have some handy context ready to be sent with your question, including highlighted passages, saved words, notes you’ve written, or the full chapter.

If you had a good chat and want to save it for later, you can do so. Like everything else, it will be linked to the chapter and level you were reading during the conversation. You can also return to the saved conversations and continue from where you left off.

Flashcards#

Words you save can also be made into flashcards to be reviewed using spaced repetition. Once a saved word has a definition, it becomes ready for review.

Each time you review a word, you choose Remembered it or Forgot it:

• The first successful review schedules the word for the next day.
• The second schedules it 6 days later.
• The third schedules it 14 days later.
• Further successful reviews schedule it every 30 days.
• Choosing Forgot it resets that progress and brings the word back after ten minutes.
• Reviewing a word before it is due does not advance its schedule.

After three successful reviews on their scheduled dates the word becomes Known. Known words still return for review so they stay fresh. The flashcard includes the word’s source sentence and book, keeping the vocabulary in the context where you found it.

Keyboard bindings and a command palette#

I’m a keyboard person, and I also have poor memory. So I added a command palette with handy shortcuts and keybindings to help me out until muscle memory takes over.

Other little things#

I love little details:

• You can bookmark pages.
• Custom dictionary websites can be configured per language so you can easily look up words without AI.
• In the section menu, you can see which adaptations are available, whether it is bookmarked, and whether there are any saved words, highlights, notes, or AI chats.
• You can see how many words are in each chapter and get an estimated reading time.
• A chapter can be read in parts, so you only get what fits in the viewport, or you can choose scroll mode to have the whole chapter at once.
• You can also select a passage and instead of highlighting it, you can send it directly to the AI, which will add it to your context without saving it.
• AI features are optional and can be disabled. The app is fully functional, and you can still use it to read normally, use system voices for narration, save words, highlights, notes, and so on.

And I almost forgot. We also have light mode ☀️

Built with#

Layer	Tech
Frontend	Vite, React, TypeScript, Tailwind, Zustand
Backend	Rust + Tauri
Storage	SQLite, Markdown, and local media files
AI	OpenAI API for adaptations, chat, and generated narration

A TypeScript/React front end talks to a Rust core over Tauri’s IPC bridge. Rust handles EPUB parsing, filesystem access, SQLite, AI requests, and credentials, while Zustand keeps only temporary UI state. Structured data such as book metadata, reading progress, flashcard scheduling, and conversations lives in SQLite. Content that remains useful outside Lingon (adaptations, notes, flashcards, narration, and recordings) is stored as open Markdown or media files in a library folder you choose.

What’s next#

• Highlighted passages as flashcards. The idea here is to give you the opportunity to try to translate passages you chose to review. There is no need to translate them directly, you could instead explain what you understood if you can’t translate something exactly, and have an LLM give you feedback on how well you understood the passage.
• A reading dashboard. I’m itching to get the dashboard ready. I want to add some traditional NLP analysis for each book and create an algorithm that estimates its complexity and reading level.