How this works

You give us a Spanish text. We split it into words, work out each word’s dictionary form, count them, and measure how the text reads. You get a frequency-ranked list, a difficulty score, and a breakdown of the verb forms the text uses.

How do I get help?

Write to support@vocabify.xyz. Most replies within two business days.

You don’t need an account to write, and if the problem is that you can’t sign in, this is the right place — say what you tried and what happened.

Why can’t I upload my Kindle book?

Kindle books are copy-protected, and removing that protection is unlawful. This is a boundary, not a backlog item.

Two things still work. The library holds analyses of books nobody needs to upload again — statistics are facts about a work, so a book being copy-protected does not stop us describing it. And pasting a chapter gives you a real sample of the vocabulary.

What do you store?

Not your file. The bytes exist for the length of one request and are gone when it returns — there is no temporary storage and no upload folder. What is kept is the derived statistics: counts, scores, and the word list.

Nothing you analyse appears in the library unless you choose to list it.

Why not just use Readlang, LingQ, or Language Reactor?

Those are reading companions — they help while you read. This is preparation and selection: deciding whether a book is worth starting, and learning its vocabulary before you do.

Where does the difficulty score come from?

The Fernández Huerta formula, a published measure of Spanish reading ease calculated from syllables per word and words per sentence. Higher is easier. We group scores into five bands so books can be compared at a glance, and every band carries its number and its name — never colour alone.

The bands are fixed, not ranked against our own catalogue. A book’s band means the same thing next year as it does today.

Why is this word wrong?

Finding a word’s dictionary form is done by a statistical model, and it is not perfect — around one word in a hundred goes wrong on ordinary prose, more on dialect, poetry, or text from a bad scan. When a text scores unusually badly we say so above the list rather than presenting it as confident.

Corrections are worth sending. A wrong entry usually means a wrong rule, and fixing one fixes it for every book. There is a Report a wrong word control at the bottom of every analysis — it attaches which analysis and which version of the engine produced the word, which is usually the difference between something we can fix and something we have to write back about first. No account needed, and nothing from your book is sent.

Why only Spanish?

Accuracy is language-specific work: the dictionary forms, the verb paradigms, the frequency data and the corrections all have to be built and checked per language. One language done properly is more useful than five done approximately.