Extracting Vocab from Books with BookToAnki (That Actually Fits Your Level)

If you actually read long-form English books, the standard vocabulary workflow has been completely broken for the last decade.

You find a brilliant word. You look it up in the dictionary. You forget it exactly three days later. The obvious, mathematical solution is Anki spacing, but physically getting a word out of an EPUB novel and into Anki is a violently tedious workflow.

Most vocabulary apps try to solve this by arrogantly throwing a generic "Top 1000 GRE Words" deck at you. The newer startups are even worse—they are just thin, lazy UI wrappers around OpenAI APIs, asking you to upload your entire DRM-free library to a random server just to get a study list back.

Generic apps violently ignore how you actually read

Memorizing alphabetical frequency lists is a miserable experience.

Language acquisition is infinitely faster when the word arrives with emotional context. If you encounter a weird adjective during a chaotic sci-fi space battle, your brain permanently attaches that word to the physical scene, the character's anger, and the tension of the chapter.

BookToAnki exists to preserve that exact tension. It doesn't just rip the word out; it aggressively grabs the exact sentence you were reading. When the Anki card pops up three weeks later, you don't just see a sterile dictionary entry—you are physically instantly transported back to the exact moment you encountered it.

Book tools should not be generic AI chatbots

I am exhausted by every single productivity app mutating into a generic generative AI pipeline.

I don't want to upload my private reading library to a consumer LLM. BookToAnki executes the vocabulary extraction entirely locally. It is completely blind to everything except your vocabulary. It cannot write poetry, it cannot generate emails, and it cannot summarize your book.

The real bottleneck is human filter

My V1 prototype of this script was a total disaster. It automatically extracted every single unknown word from a 300-page book and spat out an 800-card Anki deck. I opened it once, was physically repulsed by the volume, and never touched it again.

You mathematically cannot learn everything. You have to ruthlessly filter.

BookToAnki targets specific bands (C1, TOEFL, IELTS). But even then, the system absolutely demands a human pass. If a word is insanely rare, violently delete it. Let it stay in the book.

If you are tired of generic vocabulary apps that ignore your actual reading habit, try it out.