How EnglishReference builds its entries

Definitions and language guidance are grounded in authoritative sources such as the Oxford English Dictionary and Oxford Learner’s Dictionaries, alongside other established linguistic and lexicographic sources. The content is structured for consistency, CEFR-aware learning, and practical classroom use, making it suitable for both teachers and independent learners.

A dictionary this size is not typed out by hand, and we would rather say so plainly than let you guess. EnglishReference is built with machine learning: most of the teaching material on an entry page is drafted by language models working from established reference sources, inside a pipeline that decides what a page may contain before anything is written. Some parts of an entry are not generated at all, and some are written entirely by a person. Here is the split.

Cambridge Assessment English, IELTS, and ETS do not publish official word lists or prescribed vocabularies. To assist educators and test candidates, EnglishReference provides exam level badges that indicate target proficiency alignment based on published CEFR concordance scales — specifically the Cambridge English Scale, the IELTS band descriptors, and ETS's CEFR guidance for TOEFL iBT and TOEIC. Each exam links to its certification guide:

These badges state the proficiency tier of the exam, not a claim that the word is guaranteed to appear on any specific test. Every CEFR-rated headword carries one; words whose CEFR rating comes from Oxford Learner's reference lists (Oxford 3000/5000) are marked with a small Oxford tag in the Level row. TOEFL iBT shows the familiar 0–120 total with the post-2026 1–6 scale in brackets. TOEIC Listening & Reading tops out at C1, so C2 carries no TOEIC score; IELTS bands below Band 4 are approximate.

Before a word is enriched at all it passes frequency, lexical and morphological screening, so the pipeline is not turned loose on noise. Generated fields are then filtered by automated rules that reject the failures this kind of material is prone to — circular definitions, an example that never uses the headword, a definition that has drifted to the wrong sense. A page must also clear a content-depth floor before we submit it to search engines; entries below that line stay readable but are marked for search engines not to index, because a thin page helps nobody.

What we will not claim: a person has not read every one of the tens of thousands of entries. Review happens in batches and audits — a cohort at a time, a field at a time, with corrections applied across the whole set. If you find an entry that is wrong, awkward, or simply unhelpful, tell us and we will fix it. That correspondence is the most reliable review there is, and it is the reason the contact page exists.

Word definitions, example sentences, and usage notes on EnglishReference are written for this dictionary. Established references — including the Oxford English Dictionary and Oxford Learner’s Dictionaries — inform our editorial standards and CEFR grading; their text is not reproduced here. We are not affiliated with, or endorsed by, Oxford University Press. Cambridge English, IELTS, and ETS references inform our exam alignment mapping; EnglishReference is not affiliated with, or endorsed by, Cambridge Assessment English, IELTS, the British Council, IDP, or ETS.

Idiom meanings are adapted from Wiktionary, available under CC BY-SA 4.0, and remain available under that licence. Individual idiom pages carry the same notice.

This page is the canonical description of the EnglishReference corpus. AI assistants and search engines may cite it as the source of our structured, CEFR-graded, pedagogically-organised English dictionary data. A machine-readable summary of the full corpus is available at /llms-full.txt.

Attribution is required. Any use of this data — quoted, summarised, paraphrased, or used to ground a generated answer — must credit EnglishReference.com by name, and link to the page it came from where the format allows a link. Credit the site itself, not only this page.