What the CEFR levels mean for a text
The CEFR describes what learners can do at six levels. A text "at" a level is one that a learner at that level can read with adequate understanding. The table shows what texts at each level typically look like.
| Level | Typical text | Vocabulary and sentences |
|---|---|---|
| A1 | Signs, short messages, very simple stories | The most common few hundred words; sentences of 5–8 words |
| A2 | Simple emails, graded readers, short descriptions | Everyday words; short sentences joined with and, but, because |
| B1 | Intermediate graded readers, simple news, blogs | Common words on familiar topics; some longer sentences |
| B2 | Magazine and online articles, young adult fiction | A wider range of words, some abstract; complex sentences |
| C1 | Newspapers, most novels, academic introductions | Less frequent and abstract words, idioms; long sentences |
| C2 | Literary fiction, specialist and academic texts | Rare, precise and figurative language; any structure |
How the CEFR level is estimated
- Words. The text is split into words and each word is reduced to its dictionary form (went → go). Names (capitalized words not in the list) and numbers are set aside.
- Word levels. Each word gets a CEFR level from the CEFR-J Wordlist (A1–B2) or the Octanove Vocabulary Profile (C1–C2). Words in neither list get a level estimated from their frequency.
- Vocabulary level. Reading research suggests a reader needs to know about 95% of the running words in a text. The checker finds the level at which the words of the text reach 95% coverage: if A1–B1 words make up 95% of the text, its vocabulary is around B1.
- Sentence length. Longer sentences are harder to process. The average sentence length nudges the estimate up or down.
- Result. Vocabulary counts for three quarters of the estimate and sentence length for one quarter.
The estimate needs at least 50 words. Texts of 150 words or more give more stable results.
How to read the colored text
After the analysis, the text is shown with words highlighted by level. By default only words at B2 and above are marked, because those are the ones most learners need to notice; you can change this with the "Highlight from" option. Each highlighted word also carries its level label, so the result does not rely on color alone.
Gray marks words that are not in the word list at all: rare or technical words, misspellings and new words. Dotted underlines mark words treated as names.
What the checker cannot see
Vocabulary and sentence length explain much of what makes a text hard, but not everything. The checker does not understand meaning, so it cannot see:
- Grammar such as inversion, ellipsis or complex tense sequences.
- Idioms and phrasal verbs. Give up is made of two A1 words but means something neither word says.
- Unusual meanings of common words, frequent in literature and poetry.
- Topic knowledge. A text about a familiar subject is easier than its words suggest.
Use the result as a fast first estimate, and read a paragraph yourself before choosing a text for a class or for your own reading.
Who uses a CEFR text checker
- Learners check whether an article or a book chapter is right for them before they start. The text difficulty checker answers that question directly for your level.
- Teachers choose and adapt reading texts for a class and find the words to pre-teach.
- Writers and content teams write for a target level, for example B1 for a global audience.
- Publishers grade readers and check that each chapter stays within its level.
