The full dataset is not published yet. A license review comes first, and the dataset will be released once it is finished; no date is promised. Until then the site shows the dataset's statistics and a sample, and lists cannot be built or downloaded.

Common English OpenList

About

The Common English OpenList is made by Ryan Kamp. It grew out of the English OpenList, a word list with its own rules, which the CEOL Universe now holds as one of its parts.

Why it exists

Ask whether a string is a word, and the answer depends on which dictionary you ask. Dictionaries and word lists disagree, for several reasons:

  • Timing: They are updated at different times, and some are no longer updated at all.
  • Scope: Specialized and technical terms are in scope for some and not for others.
  • Standards: Some leave out slang or profanity on purpose.
  • Definitions: Each has its own rules for what a word is, such as whether it may hold a space, a hyphen, or an accent, and its own way of taking in a new entry, from a community of editors or from an institution's staff.

The CEOL Universe does not settle the question. For each candidate, it records which published lists hold it, which could not have held it, and how it came in, in versions fixed at a date, so that whoever settles the question can see the evidence.

The family

The CEOL Universe
The inventory: every candidate, with its provenance, in versions that never change once released.
The English OpenList
A word list with its own rules, kept here as a dated subset of the universe.
Ars Magna
An anagram finder that shares the English OpenList's dictionary.

Licenses

The data
CC BY-SA 4.0, as the terms of Wiktionary and of wordfreq require. Every version ships its sources' notices.
The code
MIT.
This site's text
CC BY-SA 4.0, except the technical report, which is CC BY 4.0.

Sources, licenses, and credits

Language models in the data

The English OpenList decided some of the entries of its valid list with a language model, Google Gemini 3 Flash Preview: through a verification pipeline that used it, or by its verdict alone. Its synthetic engine, whose code does not survive, made other strings by affixing, by chemical substitution, by scientific prefixing, and by Greek plurals; the English OpenList's page gives its author's account of the methods. The universe marks every such row in model_or_engine, so a list can keep or drop them.

Compose a list without them

Requests and contact

Requests to withhold, correct, or restore a source or a row go to . A request is acknowledged within 7 days and acted on within 14 days of its verification, and each version's change log records every withholding, as counts.

How to cite it

This version

Version 1.0, built October 25, 2026: 10,020,150rows rows. A released version never changes; a correction comes as a new version with a change log.

Versions and changes

The author

Ryan Kamp makes the Common English OpenList and its sister projects.