WordieLexical Coverage Calculator
Paste something you want to read, tap every word you don't know, and see what percentage of the text you already understand. Research suggests that number predicts whether reading will feel easy, learnable or frustrating.
What does lexical coverage measure?
Lexical coverage is the percentage of running words in a text that you know. Running words means every occurrence counts: if an unknown word appears five times, it costs you five words of coverage. If you know 490 of the 500 words in an article, your coverage is 98%. It is one of the best single predictors of whether a text will be readable for you, much better than a general readability score, because it is measured against your own vocabulary rather than an average reader's.
How the calculation works
coverage = (total running words − unknown running words) / total running words × 100
The tool also shows words to learn for 98%: it sorts your unknown words by how often they appear and counts how many you would need to learn, starting with the most frequent, to bring the text up to 98% coverage. On a long text, that number is often surprisingly small, because one or two key terms repeat throughout.
How much coverage do you need? The 95% and 98% thresholds
Hu and Nation (2000) gave learners fiction texts with different proportions of unknown words and tested their comprehension. Comprehension rose steadily with coverage, and at 98% most readers understood the text adequately without help. Nation (2006) used this figure to estimate that reading a wide range of English texts comfortably takes knowledge of roughly 8,000 to 9,000 word families. Other research has treated 95% as a lower threshold for reasonable comprehension with support.
| Coverage | Unknown words | Reading experience | Best use |
|---|---|---|---|
| 98% or more | 1 in 50 or fewer | Comfortable | Extensive reading for fluency and incidental learning |
| 95–98% | 1 in 20 to 1 in 50 | Learnable with effort | Intensive reading with a dictionary |
| Below 95% | More than 1 in 20 | Hard; meaning breaks down | Pre-teach vocabulary or choose an easier text |
To feel what 95% means: with ten to twelve words on a typical book line, it is about one unknown word every other line. At 98% it is one in every four or five lines, few enough to guess from context clues.
Limitations
- Self-rating is generous. People often mark half-known words as known. If you can't explain a word's meaning in this sentence, mark it.
- Not all words matter equally. An unknown word in the headline hurts more than one in an aside. Coverage treats them the same.
- Word forms are separate. Marking argue does not mark argued; tap each form you don't know.
- Thresholds are averages. Background knowledge, interest and grammar also affect comprehension, so treat 95% and 98% as guides, not hard limits.
Related reading
- Lexical coverage: definition
- How many words do you need to be fluent?
- Comprehensible input explained
- Vocabulary size and how it relates to coverage
How to use this tool
- Paste a passage you want to read into the box (a sample is loaded with three words marked).
- Tap or click every word you don't know in the text below the box. Tap again to unmark it.
- Read your coverage percentage and the interpretation: hard, learnable or comfortable.
- Click Copy unknown words to take your list to a dictionary, notebook or flashcard app.
Frequently asked questions
What is lexical coverage?
Lexical coverage is the percentage of running words in a text that a reader knows. If you know 196 of 200 words on a page, your coverage is 98%.
What percentage of words do you need to know to understand a text?
Research by Hu and Nation (2000) and Nation (2006) suggests about 98% coverage for comfortable unassisted reading, and around 95% as a minimum for reasonable comprehension with some effort.
Does marking a word mark every occurrence?
Yes. When you tap a word, every occurrence of the same form in the text is marked, because coverage counts running words and you either know the word or you don't.
Should I count names and numbers as known?
Usually yes. Names of people and places rarely block understanding, so leave them unmarked unless they genuinely confuse you.
Is my text uploaded anywhere?
No. Everything runs in your browser and nothing is saved.
Turn these numbers into real words
Wordie quietly builds the habit for you: save words while you read, see them highlighted everywhere, review in seconds.
Add Wordie to Chrome — it's free
