Is this text at my level?

How the level is decided

The text is split into words, each word is matched against the HSK 3.0 vocabulary lists, and the reported level is the lowest one at which a reader knowing every word up to and including it recognises 90% of the running words in the passage.

Three choices in that sentence are worth stating plainly, because different tools make them differently and the answers are not comparable:

• Words, not characters. Character-based analysers report a different and generally lower number, because characters are shared between easy and hard words. 银 is not hard; 银河系 is.
• Running words, not distinct words. A word that appears twenty times is twenty encounters. Counting it once makes common texts look harder than they read.
• 90%, not 95% or 98%. 90% is the conventional threshold for "this text belongs at this level" — it is a placement number, not a comfort number. For whether you can actually read something, the coverage against your own vocabulary is the better guide, and 95-98% is where reading gets comfortable.

Which HSK

HSK 3.0, the nine-level standard introduced in 2021, using an 11,643-entry list that covers levels 1 through 9. Tools built on HSK 2.0 report levels 1-6 over a much smaller vocabulary, so a text those tools call "above HSK 6" will often land at HSK 7 or 8 here rather than off the scale. Words outside the lists entirely are reported separately instead of being forced into a level.

What a level does not tell you

Vocabulary coverage is the part of difficulty that can be counted. Sentence length, grammar, register, how much the writer assumes you know — none of that is in this number, and all of it matters. A children's story and a legal notice can share a level and not share a reading experience.

The more useful question is usually not what level a text is but whether you can read it, which depends on your vocabulary rather than a syllabus. Once you have saved some words, the text analyzer answers that version of the question directly.