Why you recognize words in Anki but not while reading

It is a familiar frustration: the card flips, you answer it instantly, your retention stats look excellent — and then the same word stops you cold in an article a week later. Spaced repetition is doing its job, but a bare word-and-definition card trains recall in one fixed, artificial setting. You learn to recognize the prompt, not to recognize the word in the wild.

Reading is the opposite of fixed. The word arrives mid-sentence, in a tense or a sense you did not drill, surrounded by phrasing you have never seen. None of the cues your card relied on are there, so the recognition that felt solid in Anki simply does not transfer to the page.

A card built from a real sentence carries the things a definition strips away: the grammar, the collocations, the register, and your own memory of where you met the word. Those cues are exactly what your brain reaches for when the word reappears in something new, which is why a sentence-fronted card recognizes far more like real reading than a word-and-gloss pair ever can.

That is the difference Capecho is built around. Cards keep words in context — the exact sentence you met the word in — so review rehearses the word the way you actually encounter it, and recognition has somewhere to transfer to.

The reason most people drill bare words is honest: pausing to copy the surrounding sentence, paste it into a card, and tidy the formatting is tedious enough that the context quietly gets dropped. Capecho removes that tax. While you read on your Mac, a single shortcut runs macOS's built-in on-device text recognition — the same engine behind Live Text — to read the word and its sentence at the moment you press it, returning only the recognized text — the screen image never reaches Capecho, and nothing is uploaded.

What opens next is a preview you can edit before saving, not a silent grab. Fix a misread character, trim the sentence, adjust the gloss, then save — so the context that lands on the card is the context you actually wanted.

A context-rich card is only as good as your understanding of what the word means there. Capecho attaches a free, unmetered explanation to every captured word — its senses, part of speech, pronunciation, and a handoff to your system dictionary — so you grasp the word before it enters review rather than guessing at the gloss later.

When the sentence itself is the hard part, an in-context explanation can spell out the precise sense the word carries in your sentence. It's metered gently — ten a day, free, unlimited on Pro — and reaching the cap never blocks capture, the word explanation, review, or export.

None of this asks you to leave Anki. When you are ready, Capecho exports your captured, sentence-fronted cards to Anki and CSV, with a target-language column so a German deck and a Spanish deck never collide in the same import. The cards arrive already context-rich — the kind that transfer — without the manual building that made you skip context in the first place.

Capecho is a complement, not a replacement. It keeps the spaced-repetition engine you already trust and fixes what sits in front of it: how the words, and their context, get onto the cards.