{
  "api_version": "1",
  "updated": "2026-10-18",
  "latest_event": "2026-10-17",
  "rights": "© Frontier Log, all rights reserved.",
  "benchmark": {
    "slug": "tau2-bench",
    "name": "τ²-bench",
    "family": "tau-bench",
    "unit": "%",
    "higher_is_better": true,
    "status": "retired",
    "benchmark_url": "https://github.com/sierra-research/tau2-bench",
    "introduced": "2025-06-09",
    "creator": "Sierra",
    "introduced_source": "https://arxiv.org/abs/2506.07982",
    "settings": [
      {
        "setting": "telecom",
        "name": "Telecom",
        "url": "https://frontierlog.ai/benchmark/tau2-bench",
        "results": 1
      }
    ]
  },
  "total": 1,
  "results": [
    {
      "id": "claude-opus-4-6:tau2-bench:telecom:1",
      "model": "claude-opus-4-6",
      "model_name": "Claude Opus 4.6",
      "model_date": "2026-02-05",
      "event": "https://frontierlog.ai/event/2026-02-05-claude-opus-4-6",
      "benchmark": "tau2-bench",
      "benchmark_name": "τ²-bench",
      "setting": "telecom",
      "setting_name": "Telecom",
      "score": 99.3,
      "unit": "%",
      "printed": "99.3%",
      "reported_by": "anthropic",
      "label": "Opus 4.6",
      "source": "https://www.anthropic.com/news/claude-opus-4-6",
      "source_title": "Claude Opus 4.6",
      "source_date": "2026-02-05",
      "where": "benchmark table (image), row 'Agentic tool use, τ2-bench', Telecom line, column 'Opus 4.6'",
      "own": true,
      "primary": true,
      "parallel": false,
      "notes": "",
      "caveat": "",
      "human_verified": true,
      "human_verified_on": "2026-10-18",
      "source_image": "https://frontierlog.ai/sources/e6c54904fe1e4f51.png",
      "source_image_original": "https://www-cdn.anthropic.com/images/4zrzovbb/website/f9564dd2f758237bd9dbe775674c4a375aff1e8a-2600x2968.png"
    }
  ]
}