Is this book too hard for me?
Check a book before you read it: an estimated level, and the words worth learning first. Your browser reads the file; it is never uploaded.
An EPUB or a PDF with text. On a computer you can also drop the file here.
Result:
Estimated level
An estimate from word frequency, not a test.
This text is short, so the estimate is rough.
Common words in this book
Share of the book's words among the most common words, counted by dictionary form.
Words worth learning first
The words this book uses most that are not among the 3,000 most common, with an example from the book.
A text file for Anki: the word, its sentence, an empty field for the meaning, and the tag Readnslate.
Every word of this text is among the 3,000 most common.
Comfortable reading needs about 95 to 98% known words. Readnslate translates the rest with one tap.
Get Readnslate on Google PlayOn a computer: the Google Play page can install Readnslate on your Android phone. Copy the book to the phone's Downloads folder, then add it in Readnslate.
How the estimate works
The checker finds each word's dictionary form, so “comí”, “comes” and “comer” count as one word. Names and numbers are left out. It then counts how many of the book's words are among the 1,000, 2,000, 3,000, 5,000 and 8,000 most common words of its language.
Studies of reading in a second language suggest that a reader needs to know about 95% of the words to follow a text, and about 98% to read it with ease.
The level comes from the share of words among the 3,000 most common:
- 92% or more: Easier
- 88 to 92%: Intermediate
- 84 to 88%: Upper intermediate
- Below 84%: Advanced
These percentages are for Spanish. Each other language has its own, 1 to 8 points lower, because the same simple story and the Universal Declaration of Human Rights score lower in those languages: 1 point for English, French, Italian and Portuguese, 2 for Dutch, 3 for German, 4 for Turkish, 7 for Polish and 8 for Finnish.
These are not CEFR levels such as B1 or C1. Word frequency says nothing about grammar, sentence length, style or the subject, so the result is a first guide.
Where it is weaker
- Old books: many old spellings count as today's words, but old words count as rare ones, so a book from the 1800s usually looks harder than a modern one. Polish, Finnish and Turkish were tested only with such old books.
- Old books in plain words, such as Don Quijote, can look easier than they are: word counts don't see grammar or style.
- Poetry, plays, and texts of a few hundred words.
- Finnish and Turkish add many endings to a word. Common endings are taken off, but some forms stay unknown, so these books can look harder than they are.
- German, Dutch and Finnish compounds count as known when their parts are.
- Names are a guess: a capitalized word that never appears in lower case is left out. German writes every noun with a capital letter, so there the guess is rougher.
Your file stays on your device
Your browser reads the file. Nothing is uploaded, and the page sets no cookies. When it needs them, the page loads its own word lists and PDF reader from readnslate.com. Closing the page clears the result.
Sources
Word frequencies: wordfreq 3.2.0 by Robyn Speer, CC BY-SA 4.0, built from Wikipedia, OpenSubtitles (opensubtitles.org), SUBTLEX word lists by Marc Brysbaert and colleagues (freely available data), Google Books Ngrams and other sources. Word forms: Wiktionary contributors, CC BY-SA 4.0, extracted by kaikki.org. Words with two possible dictionary forms are shared out using Universal Dependencies treebanks (CC BY 4.0, CC BY-SA 3.0 and 4.0). The word lists on this page adapt these sources and are shared under CC BY-SA 4.0; full list of sources. PDFs are read with PDF.js by Mozilla (Apache License 2.0).