{
  "id": "webtext",
  "name": "WebText",
  "builder": "OpenAI",
  "release_date": {
    "value": null,
    "status": "not_recorded"
  },
  "availability": {
    "value": "partial",
    "checked": "2026-09-24",
    "source": "https://github.com/openai/gpt-2-output-dataset",
    "note": "OpenAI released 250K documents from the WebText test set; the training set was not released."
  },
  "content": {
    "value": "Millions of web pages linked from Reddit; the GPT-2 training corpus.",
    "source": "https://cdn.openai.com/better-language-models/language_models_are_unsupervised_multitask_learners.pdf",
    "status": "recorded",
    "note": "The gpt2-xl record carries the paper's fuller description (~8M documents, outbound links with karma >= 3)."
  },
  "primary_sources": [
    "https://cdn.openai.com/better-language-models/language_models_are_unsupervised_multitask_learners.pdf",
    "https://github.com/openai/gpt-2-output-dataset"
  ],
  "record_history": [
    {
      "date": "2026-09-24",
      "change": "created from primary sources (dataset records ruling)",
      "by": "wilson-pruitt + claude"
    }
  ]
}