{
  "id": "gpt-3",
  "name": "GPT-3 (175B, 2020 paper)",
  "developer": "OpenAI",
  "release_date": {
    "value": "2020-05-28",
    "source": "https://arxiv.org/abs/2005.14165",
    "status": "partial",
    "note": "arXiv v1 date of the paper."
  },
  "weights_status": "api_only",
  "availability": {
    "value": "never_released",
    "checked": "2026-09-24",
    "source": "https://arxiv.org/abs/2005.14165",
    "note": "Weights never released; API access status not tracked by this field."
  },
  "license": {
    "value": "Proprietary; no weights released.",
    "source": "https://arxiv.org/abs/2005.14165",
    "status": "partial"
  },
  "architecture": {
    "family": "decoder_only",
    "note": "Stub specimen, recorded because five open models name GPT-3's paper as their design reference. Paper sec. 2.1: 'We use the same model and architecture as GPT-2 ... with the exception that we use alternating dense and locally banded sparse attention patterns.' The paper describes 8 sizes; values are for 175B (Table 2.1).",
    "n_layers": {
      "value": 96,
      "source": "https://arxiv.org/abs/2005.14165",
      "status": "recorded"
    },
    "hidden_size": {
      "value": 12288,
      "source": "https://arxiv.org/abs/2005.14165",
      "status": "recorded"
    },
    "n_heads": {
      "value": 96,
      "source": "https://arxiv.org/abs/2005.14165",
      "status": "recorded"
    },
    "vocab_size": {
      "value": null,
      "status": "not_recorded"
    },
    "positional_encoding": {
      "value": null,
      "status": "not_recorded"
    },
    "context_length": {
      "value": 2048,
      "source": "https://arxiv.org/abs/2005.14165",
      "status": "recorded"
    }
  },
  "training_data": {
    "value": "300B tokens from a weighted mix of filtered Common Crawl, WebText2, Books1, Books2 and English Wikipedia.",
    "source": "https://arxiv.org/abs/2005.14165",
    "status": "recorded"
  },
  "techniques": [
    "transformer-decoder",
    "byte-pair-encoding"
  ],
  "primary_sources": [
    "https://arxiv.org/abs/2005.14165"
  ],
  "record_history": [
    {
      "date": "2026-09-24",
      "change": "created as a stub specimen to resolve design_follows edges",
      "by": "wilson-pruitt + claude"
    }
  ]
}