Downloads

From the dataset

• The list

  data/merged_valid_words.txt: 345,162 words, about 4 MB.

• Every word's record

  data/merged_valid_dict.json: how each word got on the list, about 280 MB. Fields vary by how the word came in; the dataset's card describes them.

• The candidates

  data/merged_invalid_words.txt: 9,308,837 strings not on the list, about 90 MB. The nightly run checks about 1,000 of them a night.

• Word-game files

  latest/brrrdle/: answers and allowed guesses for lengths 2 to 35, in JSON.

Subsets

• Without the synthetic engine

  without-synthetic.txt: 313,919 words, 3.4 MB. Every entry except those the synthetic engine made.

• Lengths 2 to 45

  lengths-2-to-45.txt: 345,157 words, 3.8 MB. The bounds of tournament word lists. The full list breaks both on purpose: it holds 1-letter words and words longer than 45 letters.

Load it in code

Each prints the number of words: 345,162 for the list this site was built from.

from huggingface_hub import hf_hub_download

path = hf_hub_download(
    "ryanjosephkamp/english-openlist",
    "data/merged_valid_words.txt",
    repo_type="dataset",
)
words = open(path, encoding="utf-8").read().split()
print(len(words))

const url =
  "https://huggingface.co/datasets/ryanjosephkamp/english-openlist/resolve/main/data/merged_valid_words.txt";
const words = (await (await fetch(url)).text()).split("\n").filter(Boolean);
console.log(words.length);

curl -sL -o merged_valid_words.txt \
  https://huggingface.co/datasets/ryanjosephkamp/english-openlist/resolve/main/data/merged_valid_words.txt
wc -l < merged_valid_words.txt