{
  "conclusion": {
    "cache_continuation_regression_fixed": true,
    "cost_regression_vs_frozen_current_codex_fixed_in_all_tasks": true,
    "speed_regression_vs_frozen_current_codex_fixed_in_all_tasks": false,
    "remaining_issue": "QueryForge median wall time was 12.99% slower than the frozen current-Codex control."
  },
  "fixed_lane": {
    "captured_requests": 169,
    "continuation_drops": 0,
    "midrun_commentary_trigger_runs": [
      "queuecraft/wave2",
      "queuecraft/wave5"
    ],
    "passes": 15,
    "runs": 15,
    "spend_usd": 7.2407715,
    "synthetic_continue_requests": 0,
    "weighted_cache_share": 0.8702381636421608
  },
  "frozen_control": {
    "codex_commit": "db75c19352d29ef29c17dbcf73a7244f1b1a8d10",
    "codex_version": "0.144.1",
    "runs_per_task": 5
  },
  "model": "gpt-5.6-sol",
  "pricing_usd_per_1m_tokens": {
    "cached_input": 0.5,
    "input": 5.0,
    "output": 30.0
  },
  "reasoning_effort": "high",
  "robustness_runs_excluded_from_primary_comparison": {
    "captured_requests": 29,
    "continuation_drops": 0,
    "passes": 3,
    "runs": [
      "queuecraft/wave6",
      "queuecraft/wave7",
      "queuecraft/wave8"
    ],
    "spend_usd": 1.460259,
    "synthetic_continue_requests": 0
  },
  "service_tier": "default",
  "tasks": {
    "queryforge": {
      "captured_requests": 56,
      "continuation_drops": 0,
      "fixed": {
        "median_cost_usd": 0.4778395,
        "median_seconds": 154.07,
        "passes": 5,
        "runs": 5,
        "weighted_cache_share": 0.8777816361926095
      },
      "frozen_current_codex": {
        "fixed_cost_delta_pct": -7.766613617873053,
        "fixed_time_delta_pct": 12.986851079120875,
        "median_cost_usd": 0.5180765,
        "median_seconds": 136.361
      },
      "synthetic_continue_requests": 0
    },
    "queuecraft": {
      "captured_requests": 46,
      "continuation_drops": 0,
      "fixed": {
        "median_cost_usd": 0.4144095,
        "median_seconds": 53.677,
        "passes": 5,
        "runs": 5,
        "weighted_cache_share": 0.8300423070718815
      },
      "frozen_current_codex": {
        "fixed_cost_delta_pct": -8.500998539449768,
        "fixed_time_delta_pct": -34.99134057576087,
        "median_cost_usd": 0.4529115,
        "median_seconds": 82.569
      },
      "synthetic_continue_requests": 0
    },
    "textwright": {
      "captured_requests": 67,
      "continuation_drops": 0,
      "fixed": {
        "median_cost_usd": 0.4940225,
        "median_seconds": 70.474,
        "passes": 5,
        "runs": 5,
        "weighted_cache_share": 0.8909713149619457
      },
      "frozen_current_codex": {
        "fixed_cost_delta_pct": -5.276110438423176,
        "fixed_time_delta_pct": -5.388787455697552,
        "median_cost_usd": 0.5215395,
        "median_seconds": 74.488
      },
      "synthetic_continue_requests": 0
    }
  }
}
