{
  "schemaVersion": 1,
  "resultVersion": "v1",
  "generatedBy": "scripts/build-backrooms-results-table.mjs",
  "generatedAt": "2026-09-21T20:33:37.065Z",
  "dataUpdatedAt": "2026-09-20",
  "canonicalUrl": "https://benchmarks.bio/results/v1/",
  "methodology": "Overall score is the equally weighted mean of observed benchmark scores.",
  "defaultScope": {
    "assessment": "capabilities",
    "aggregation": "Overall score is the equally weighted mean of observed benchmark scores.",
    "missingData": "Missing benchmark scores are excluded, never treated as zero"
  },
  "benchmarks": [
    {
      "id": "spatialbench",
      "label": "SpatialBench",
      "category": "multiomics",
      "horizon": "short",
      "evalCount": 115,
      "scoreScope": {
        "label": "Verified subset",
        "evaluationCount": 115,
        "comparisonSet": {
          "label": "Full benchmark",
          "evaluationCount": 159
        }
      },
      "weight": 1
    },
    {
      "id": "spatialbench-long",
      "label": "SpatialBench-Long",
      "category": "multiomics",
      "horizon": "long",
      "evalCount": 24,
      "scoreScope": {
        "label": "Verified subset",
        "evaluationCount": 22,
        "comparisonSet": {
          "label": "Full benchmark",
          "evaluationCount": 24
        }
      },
      "weight": 1
    },
    {
      "id": "scbench",
      "label": "scBench",
      "category": "multiomics",
      "horizon": "short",
      "evalCount": 195,
      "scoreScope": {
        "label": "Full benchmark",
        "evaluationCount": 195
      },
      "weight": 1
    },
    {
      "id": "scbench-long",
      "label": "scBench-Long",
      "category": "multiomics",
      "horizon": "long",
      "evalCount": 21,
      "scoreScope": {
        "label": "Verified merged subset",
        "evaluationCount": 22,
        "comparisonSet": {
          "label": "Original full release",
          "evaluationCount": 21
        }
      },
      "weight": 1
    },
    {
      "id": "epibench",
      "label": "EpiBench",
      "category": "multiomics",
      "horizon": "short",
      "evalCount": 106,
      "scoreScope": {
        "label": "Full benchmark",
        "evaluationCount": 106
      },
      "weight": 1
    },
    {
      "id": "variantbench",
      "label": "VariantBench",
      "category": "multiomics",
      "horizon": "short",
      "evalCount": 118,
      "scoreScope": {
        "label": "Full benchmark",
        "evaluationCount": 118
      },
      "weight": 1
    },
    {
      "id": "txbench-ab",
      "label": "TxBench-Antibody-Discovery",
      "category": "therapeutics",
      "horizon": "short",
      "evalCount": 100,
      "scoreScope": {
        "label": "Full benchmark",
        "evaluationCount": 100
      },
      "weight": 1
    },
    {
      "id": "txbench-pp",
      "label": "TxBench-Preclinical-Pharmacology",
      "category": "therapeutics",
      "horizon": "short",
      "evalCount": 100,
      "scoreScope": {
        "label": "Full benchmark",
        "evaluationCount": 100
      },
      "weight": 1
    },
    {
      "id": "txbench-od",
      "label": "TxBench-Oligo-Discovery",
      "category": "therapeutics",
      "horizon": "short",
      "evalCount": 113,
      "scoreScope": {
        "label": "Full benchmark",
        "evaluationCount": 113
      },
      "weight": 1
    },
    {
      "id": "biosecbench-surveillance",
      "label": "BioSecBench-Surveillance",
      "category": "biosecurity",
      "horizon": "short",
      "evalCount": 102,
      "scoreScope": {
        "label": "Full benchmark",
        "evaluationCount": 102
      },
      "weight": 1
    },
    {
      "id": "biosecbench-refusal",
      "label": "BioSecBench-Refusal",
      "category": "biosecurity",
      "horizon": "short",
      "assessment": "safeguards",
      "evalCount": 107,
      "scoreScope": {
        "label": "Full benchmark",
        "evaluationCount": 107
      },
      "weight": 1
    },
    {
      "id": "functionbench",
      "label": "BioSecBench-Function",
      "category": "biosecurity",
      "horizon": "short",
      "evalCount": 111,
      "scoreScope": {
        "label": "Full benchmark",
        "evaluationCount": 111
      },
      "weight": 1
    }
  ],
  "models": [
    {
      "id": "claude-opus-4-7-claude-code",
      "model": "Claude Opus 4.7",
      "harness": "Claude Code",
      "provider": "Anthropic",
      "overallScorePct": 39.8,
      "observedBenchmarkCount": 7,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 64.06,
        "spatialbench-long": null,
        "scbench": 54.02,
        "scbench-long": 10.61,
        "epibench": 31.45,
        "variantbench": 35.6,
        "txbench-ab": 39.03,
        "txbench-pp": 44,
        "txbench-od": null,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 33.9,
        "functionbench": null
      }
    },
    {
      "id": "claude-opus-4-7-pi",
      "model": "Claude Opus 4.7",
      "harness": "Pi",
      "provider": "Anthropic",
      "overallScorePct": 47,
      "observedBenchmarkCount": 5,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 63.77,
        "spatialbench-long": null,
        "scbench": 53.94,
        "scbench-long": null,
        "epibench": 31.13,
        "variantbench": 37,
        "txbench-ab": null,
        "txbench-pp": 49.33,
        "txbench-od": null,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 36.7,
        "functionbench": null
      }
    },
    {
      "id": "claude-opus-4-8-claude-code",
      "model": "Claude Opus 4.8",
      "harness": "Claude Code",
      "provider": "Anthropic",
      "overallScorePct": 42.8,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 68.99,
        "spatialbench-long": 28.79,
        "scbench": 57.95,
        "scbench-long": 21.21,
        "epibench": 37.42,
        "variantbench": 39.8,
        "txbench-ab": 42.04,
        "txbench-pp": 54.67,
        "txbench-od": 39.88,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 26.1,
        "functionbench": 37
      }
    },
    {
      "id": "claude-opus-4-8-pi",
      "model": "Claude Opus 4.8",
      "harness": "Pi",
      "provider": "Anthropic",
      "overallScorePct": 44.8,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 66.38,
        "spatialbench-long": 28.79,
        "scbench": 57.29,
        "scbench-long": 25.76,
        "epibench": 38.99,
        "variantbench": 42.1,
        "txbench-ab": 45.64,
        "txbench-pp": 59.33,
        "txbench-od": 39.4,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 28.6,
        "functionbench": 44.6
      }
    },
    {
      "id": "claude-opus-5-claude-code",
      "model": "Claude Opus 5",
      "harness": "Claude Code",
      "provider": "Anthropic",
      "overallScorePct": 48.6,
      "observedBenchmarkCount": 11,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 71.01,
        "spatialbench-long": 28.79,
        "scbench": 60.14,
        "scbench-long": 30.3,
        "epibench": 31.13,
        "variantbench": 49.72,
        "txbench-ab": 52.96,
        "txbench-pp": 68.67,
        "txbench-od": 48.08,
        "biosecbench-surveillance": 43.6,
        "biosecbench-refusal": 28.9,
        "functionbench": 50.3
      }
    },
    {
      "id": "claude-opus-5-pi",
      "model": "Claude Opus 5",
      "harness": "Pi",
      "provider": "Anthropic",
      "overallScorePct": 48.2,
      "observedBenchmarkCount": 11,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 66.67,
        "spatialbench-long": 31.82,
        "scbench": 59.49,
        "scbench-long": 36.36,
        "epibench": 35.22,
        "variantbench": 49.44,
        "txbench-ab": 51.39,
        "txbench-pp": 63.33,
        "txbench-od": 42.18,
        "biosecbench-surveillance": 51.5,
        "biosecbench-refusal": 29.4,
        "functionbench": 43
      }
    },
    {
      "id": "claude-sonnet-5-claude-code",
      "model": "Claude Sonnet 5",
      "harness": "Claude Code",
      "provider": "Anthropic",
      "overallScorePct": 39.1,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 64.35,
        "spatialbench-long": 15.15,
        "scbench": 56.7,
        "scbench-long": 16.67,
        "epibench": 37.74,
        "variantbench": 38.7,
        "txbench-ab": 37.26,
        "txbench-pp": 48.33,
        "txbench-od": 39.29,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 27.4,
        "functionbench": 37.2
      }
    },
    {
      "id": "claude-sonnet-5-pi",
      "model": "Claude Sonnet 5",
      "harness": "Pi",
      "provider": "Anthropic",
      "overallScorePct": 39.5,
      "observedBenchmarkCount": 9,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 69.28,
        "spatialbench-long": 24.24,
        "scbench": 59.52,
        "scbench-long": 19.7,
        "epibench": 35.22,
        "variantbench": 38.1,
        "txbench-ab": 33.86,
        "txbench-pp": 42,
        "txbench-od": null,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 30.1,
        "functionbench": 33.8
      }
    },
    {
      "id": "grok-4-5-pi",
      "model": "Grok 4.5",
      "harness": "Pi",
      "provider": "SpaceXAI",
      "overallScorePct": 41.3,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 71.01,
        "spatialbench-long": 22.73,
        "scbench": 56.58,
        "scbench-long": 25.76,
        "epibench": 28.93,
        "variantbench": 37.6,
        "txbench-ab": 38.85,
        "txbench-pp": 57.67,
        "txbench-od": 38.35,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 9.3,
        "functionbench": 35.6
      }
    },
    {
      "id": "grok-4-6-pi",
      "model": "Grok 4.6",
      "harness": "Pi",
      "provider": "SpaceXAI",
      "overallScorePct": 44.8,
      "observedBenchmarkCount": 11,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 69.86,
        "spatialbench-long": 19.7,
        "scbench": 59.66,
        "scbench-long": 24.24,
        "epibench": 34.28,
        "variantbench": 38.1,
        "txbench-ab": 51.85,
        "txbench-pp": 62,
        "txbench-od": 43.95,
        "biosecbench-surveillance": 47.5,
        "biosecbench-refusal": 63,
        "functionbench": 42
      }
    },
    {
      "id": "grok-4-6-grok-build",
      "model": "Grok 4.6",
      "harness": "Grok Build",
      "provider": "SpaceXAI",
      "overallScorePct": 48.2,
      "observedBenchmarkCount": 11,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 77.39,
        "spatialbench-long": 30.3,
        "scbench": 61.54,
        "scbench-long": 27.27,
        "epibench": 37.42,
        "variantbench": 46,
        "txbench-ab": 51,
        "txbench-pp": 64.67,
        "txbench-od": 42.77,
        "biosecbench-surveillance": 48,
        "biosecbench-refusal": 59,
        "functionbench": 44.1
      }
    },
    {
      "id": "gemini-3-5-flash-pi",
      "model": "Gemini 3.5 Flash",
      "harness": "Pi",
      "provider": "Google",
      "overallScorePct": 41.7,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 66.96,
        "spatialbench-long": 30.3,
        "scbench": 56.92,
        "scbench-long": 24.24,
        "epibench": 28.93,
        "variantbench": 35.3,
        "txbench-ab": 42.62,
        "txbench-pp": 51.33,
        "txbench-od": 38.64,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 51.5,
        "functionbench": 42.2
      }
    },
    {
      "id": "gpt-5-5-codex",
      "model": "GPT-5.5",
      "harness": "Codex",
      "provider": "OpenAI",
      "overallScorePct": 38.7,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 64.93,
        "spatialbench-long": 33.33,
        "scbench": 57.78,
        "scbench-long": 16.67,
        "epibench": 39.94,
        "variantbench": 28.8,
        "txbench-ab": 31.53,
        "txbench-pp": 47.33,
        "txbench-od": 30.38,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 26.7,
        "functionbench": 36.5
      }
    },
    {
      "id": "gpt-5-5-pi",
      "model": "GPT-5.5",
      "harness": "Pi",
      "provider": "OpenAI",
      "overallScorePct": 40.1,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 65.22,
        "spatialbench-long": 30.3,
        "scbench": 56.68,
        "scbench-long": 22.73,
        "epibench": 44.97,
        "variantbench": 29.7,
        "txbench-ab": 33.56,
        "txbench-pp": 55.33,
        "txbench-od": 33.63,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 37.1,
        "functionbench": 29.1
      }
    },
    {
      "id": "gpt-5-6-luna-codex",
      "model": "GPT-5.6 Luna",
      "harness": "Codex",
      "provider": "OpenAI",
      "overallScorePct": 32.2,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 60.58,
        "spatialbench-long": 27.27,
        "scbench": 49.23,
        "scbench-long": 12.12,
        "epibench": 31.45,
        "variantbench": 29.9,
        "txbench-ab": 18.24,
        "txbench-pp": 42,
        "txbench-od": 23.89,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 33.3,
        "functionbench": 26.9
      }
    },
    {
      "id": "gpt-5-6-luna-pi",
      "model": "GPT-5.6 Luna",
      "harness": "Pi",
      "provider": "OpenAI",
      "overallScorePct": 33.8,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 57.1,
        "spatialbench-long": 21.21,
        "scbench": 51.9,
        "scbench-long": 28.79,
        "epibench": 36.48,
        "variantbench": 28.2,
        "txbench-ab": 19.39,
        "txbench-pp": 46.67,
        "txbench-od": 23.3,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 31.5,
        "functionbench": 25.4
      }
    },
    {
      "id": "gpt-5-6-sol-codex",
      "model": "GPT-5.6 Sol",
      "harness": "Codex",
      "provider": "OpenAI",
      "overallScorePct": 45.9,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 74.49,
        "spatialbench-long": 36.36,
        "scbench": 60.34,
        "scbench-long": 43.94,
        "epibench": 42.14,
        "variantbench": 42.1,
        "txbench-ab": 33.33,
        "txbench-pp": 59,
        "txbench-od": 34.51,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 38.2,
        "functionbench": 32.3
      }
    },
    {
      "id": "gpt-5-6-sol-pi",
      "model": "GPT-5.6 Sol",
      "harness": "Pi",
      "provider": "OpenAI",
      "overallScorePct": 47.5,
      "observedBenchmarkCount": 9,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 73.04,
        "spatialbench-long": 36.36,
        "scbench": 62.13,
        "scbench-long": 40.91,
        "epibench": 39.94,
        "variantbench": 41.2,
        "txbench-ab": 33.78,
        "txbench-pp": 61.67,
        "txbench-od": null,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 39.5,
        "functionbench": 38.3
      }
    },
    {
      "id": "gpt-5-6-terra-codex",
      "model": "GPT-5.6 Terra",
      "harness": "Codex",
      "provider": "OpenAI",
      "overallScorePct": 41.5,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 69.57,
        "spatialbench-long": 37.88,
        "scbench": 55.04,
        "scbench-long": 31.82,
        "epibench": 39.31,
        "variantbench": 38.7,
        "txbench-ab": 29.73,
        "txbench-pp": 51.67,
        "txbench-od": 32.45,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 32.3,
        "functionbench": 29.1
      }
    },
    {
      "id": "gpt-5-6-terra-pi",
      "model": "GPT-5.6 Terra",
      "harness": "Pi",
      "provider": "OpenAI",
      "overallScorePct": 42.4,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 70.43,
        "spatialbench-long": 30.3,
        "scbench": 62,
        "scbench-long": 34.85,
        "epibench": 39.94,
        "variantbench": 34.7,
        "txbench-ab": 30.98,
        "txbench-pp": 54,
        "txbench-od": 33.33,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 34.3,
        "functionbench": 33.5
      }
    },
    {
      "id": "gpt-6-astra-codex",
      "model": "GPT-6 Astra",
      "harness": "Codex",
      "provider": "OpenAI",
      "overallScorePct": 51.1,
      "observedBenchmarkCount": 11,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 70.14,
        "spatialbench-long": 34.85,
        "scbench": 64.79,
        "scbench-long": 42.42,
        "epibench": 44.65,
        "variantbench": 48.6,
        "txbench-ab": 55.22,
        "txbench-pp": 61.67,
        "txbench-od": 55.46,
        "biosecbench-surveillance": 35.2,
        "biosecbench-refusal": 23.3,
        "functionbench": 48.6
      }
    },
    {
      "id": "gpt-6-astra-pi",
      "model": "GPT-6 Astra",
      "harness": "Pi",
      "provider": "OpenAI",
      "overallScorePct": 51.3,
      "observedBenchmarkCount": 11,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 70.14,
        "spatialbench-long": 34.85,
        "scbench": 64.44,
        "scbench-long": 39.39,
        "epibench": 41.82,
        "variantbench": 47.7,
        "txbench-ab": 56.9,
        "txbench-pp": 60.33,
        "txbench-od": 53.39,
        "biosecbench-surveillance": 48.5,
        "biosecbench-refusal": 44.6,
        "functionbench": 47.1
      }
    },
    {
      "id": "kimi-k3-pi",
      "model": "Kimi K3",
      "harness": "Pi",
      "provider": "Moonshot",
      "overallScorePct": 38.1,
      "observedBenchmarkCount": 7,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 61.45,
        "spatialbench-long": 28.79,
        "scbench": 55.73,
        "scbench-long": 16.67,
        "epibench": 26.73,
        "variantbench": 32.8,
        "txbench-ab": null,
        "txbench-pp": 44.67,
        "txbench-od": null,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 8.6,
        "functionbench": null
      }
    },
    {
      "id": "deepseek-v4-1-flash-pi",
      "model": "DeepSeek V4.1 Flash",
      "harness": "Pi",
      "provider": "DeepSeek",
      "overallScorePct": 45,
      "observedBenchmarkCount": 9,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 75.36,
        "spatialbench-long": 34.85,
        "scbench": 59.49,
        "scbench-long": 27.27,
        "epibench": 42.14,
        "variantbench": null,
        "txbench-ab": 39,
        "txbench-pp": 54.33,
        "txbench-od": 33.92,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": null,
        "functionbench": 38.7
      }
    }
  ],
  "leaderboard": [
    {
      "rank": 1,
      "id": "gpt-6-astra-pi",
      "model": "GPT-6 Astra",
      "harness": "Pi",
      "provider": "OpenAI",
      "overallScorePct": 51.3,
      "observedBenchmarkCount": 11,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 70.14,
        "spatialbench-long": 34.85,
        "scbench": 64.44,
        "scbench-long": 39.39,
        "epibench": 41.82,
        "variantbench": 47.7,
        "txbench-ab": 56.9,
        "txbench-pp": 60.33,
        "txbench-od": 53.39,
        "biosecbench-surveillance": 48.5,
        "biosecbench-refusal": 44.6,
        "functionbench": 47.1
      }
    },
    {
      "rank": 2,
      "id": "gpt-6-astra-codex",
      "model": "GPT-6 Astra",
      "harness": "Codex",
      "provider": "OpenAI",
      "overallScorePct": 51.1,
      "observedBenchmarkCount": 11,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 70.14,
        "spatialbench-long": 34.85,
        "scbench": 64.79,
        "scbench-long": 42.42,
        "epibench": 44.65,
        "variantbench": 48.6,
        "txbench-ab": 55.22,
        "txbench-pp": 61.67,
        "txbench-od": 55.46,
        "biosecbench-surveillance": 35.2,
        "biosecbench-refusal": 23.3,
        "functionbench": 48.6
      }
    },
    {
      "rank": 3,
      "id": "claude-opus-5-claude-code",
      "model": "Claude Opus 5",
      "harness": "Claude Code",
      "provider": "Anthropic",
      "overallScorePct": 48.6,
      "observedBenchmarkCount": 11,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 71.01,
        "spatialbench-long": 28.79,
        "scbench": 60.14,
        "scbench-long": 30.3,
        "epibench": 31.13,
        "variantbench": 49.72,
        "txbench-ab": 52.96,
        "txbench-pp": 68.67,
        "txbench-od": 48.08,
        "biosecbench-surveillance": 43.6,
        "biosecbench-refusal": 28.9,
        "functionbench": 50.3
      }
    },
    {
      "rank": 4,
      "id": "claude-opus-5-pi",
      "model": "Claude Opus 5",
      "harness": "Pi",
      "provider": "Anthropic",
      "overallScorePct": 48.2,
      "observedBenchmarkCount": 11,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 66.67,
        "spatialbench-long": 31.82,
        "scbench": 59.49,
        "scbench-long": 36.36,
        "epibench": 35.22,
        "variantbench": 49.44,
        "txbench-ab": 51.39,
        "txbench-pp": 63.33,
        "txbench-od": 42.18,
        "biosecbench-surveillance": 51.5,
        "biosecbench-refusal": 29.4,
        "functionbench": 43
      }
    },
    {
      "rank": 5,
      "id": "grok-4-6-grok-build",
      "model": "Grok 4.6",
      "harness": "Grok Build",
      "provider": "SpaceXAI",
      "overallScorePct": 48.2,
      "observedBenchmarkCount": 11,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 77.39,
        "spatialbench-long": 30.3,
        "scbench": 61.54,
        "scbench-long": 27.27,
        "epibench": 37.42,
        "variantbench": 46,
        "txbench-ab": 51,
        "txbench-pp": 64.67,
        "txbench-od": 42.77,
        "biosecbench-surveillance": 48,
        "biosecbench-refusal": 59,
        "functionbench": 44.1
      }
    },
    {
      "rank": 6,
      "id": "gpt-5-6-sol-pi",
      "model": "GPT-5.6 Sol",
      "harness": "Pi",
      "provider": "OpenAI",
      "overallScorePct": 47.5,
      "observedBenchmarkCount": 9,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 73.04,
        "spatialbench-long": 36.36,
        "scbench": 62.13,
        "scbench-long": 40.91,
        "epibench": 39.94,
        "variantbench": 41.2,
        "txbench-ab": 33.78,
        "txbench-pp": 61.67,
        "txbench-od": null,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 39.5,
        "functionbench": 38.3
      }
    },
    {
      "rank": 7,
      "id": "gpt-5-6-sol-codex",
      "model": "GPT-5.6 Sol",
      "harness": "Codex",
      "provider": "OpenAI",
      "overallScorePct": 45.9,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 74.49,
        "spatialbench-long": 36.36,
        "scbench": 60.34,
        "scbench-long": 43.94,
        "epibench": 42.14,
        "variantbench": 42.1,
        "txbench-ab": 33.33,
        "txbench-pp": 59,
        "txbench-od": 34.51,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 38.2,
        "functionbench": 32.3
      }
    },
    {
      "rank": 8,
      "id": "deepseek-v4-1-flash-pi",
      "model": "DeepSeek V4.1 Flash",
      "harness": "Pi",
      "provider": "DeepSeek",
      "overallScorePct": 45,
      "observedBenchmarkCount": 9,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 75.36,
        "spatialbench-long": 34.85,
        "scbench": 59.49,
        "scbench-long": 27.27,
        "epibench": 42.14,
        "variantbench": null,
        "txbench-ab": 39,
        "txbench-pp": 54.33,
        "txbench-od": 33.92,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": null,
        "functionbench": 38.7
      }
    },
    {
      "rank": 9,
      "id": "grok-4-6-pi",
      "model": "Grok 4.6",
      "harness": "Pi",
      "provider": "SpaceXAI",
      "overallScorePct": 44.8,
      "observedBenchmarkCount": 11,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 69.86,
        "spatialbench-long": 19.7,
        "scbench": 59.66,
        "scbench-long": 24.24,
        "epibench": 34.28,
        "variantbench": 38.1,
        "txbench-ab": 51.85,
        "txbench-pp": 62,
        "txbench-od": 43.95,
        "biosecbench-surveillance": 47.5,
        "biosecbench-refusal": 63,
        "functionbench": 42
      }
    },
    {
      "rank": 10,
      "id": "claude-opus-4-8-pi",
      "model": "Claude Opus 4.8",
      "harness": "Pi",
      "provider": "Anthropic",
      "overallScorePct": 44.8,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 66.38,
        "spatialbench-long": 28.79,
        "scbench": 57.29,
        "scbench-long": 25.76,
        "epibench": 38.99,
        "variantbench": 42.1,
        "txbench-ab": 45.64,
        "txbench-pp": 59.33,
        "txbench-od": 39.4,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 28.6,
        "functionbench": 44.6
      }
    },
    {
      "rank": 11,
      "id": "claude-opus-4-8-claude-code",
      "model": "Claude Opus 4.8",
      "harness": "Claude Code",
      "provider": "Anthropic",
      "overallScorePct": 42.8,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 68.99,
        "spatialbench-long": 28.79,
        "scbench": 57.95,
        "scbench-long": 21.21,
        "epibench": 37.42,
        "variantbench": 39.8,
        "txbench-ab": 42.04,
        "txbench-pp": 54.67,
        "txbench-od": 39.88,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 26.1,
        "functionbench": 37
      }
    },
    {
      "rank": 12,
      "id": "gpt-5-6-terra-pi",
      "model": "GPT-5.6 Terra",
      "harness": "Pi",
      "provider": "OpenAI",
      "overallScorePct": 42.4,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 70.43,
        "spatialbench-long": 30.3,
        "scbench": 62,
        "scbench-long": 34.85,
        "epibench": 39.94,
        "variantbench": 34.7,
        "txbench-ab": 30.98,
        "txbench-pp": 54,
        "txbench-od": 33.33,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 34.3,
        "functionbench": 33.5
      }
    },
    {
      "rank": 13,
      "id": "gemini-3-5-flash-pi",
      "model": "Gemini 3.5 Flash",
      "harness": "Pi",
      "provider": "Google",
      "overallScorePct": 41.7,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 66.96,
        "spatialbench-long": 30.3,
        "scbench": 56.92,
        "scbench-long": 24.24,
        "epibench": 28.93,
        "variantbench": 35.3,
        "txbench-ab": 42.62,
        "txbench-pp": 51.33,
        "txbench-od": 38.64,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 51.5,
        "functionbench": 42.2
      }
    },
    {
      "rank": 14,
      "id": "gpt-5-6-terra-codex",
      "model": "GPT-5.6 Terra",
      "harness": "Codex",
      "provider": "OpenAI",
      "overallScorePct": 41.5,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 69.57,
        "spatialbench-long": 37.88,
        "scbench": 55.04,
        "scbench-long": 31.82,
        "epibench": 39.31,
        "variantbench": 38.7,
        "txbench-ab": 29.73,
        "txbench-pp": 51.67,
        "txbench-od": 32.45,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 32.3,
        "functionbench": 29.1
      }
    },
    {
      "rank": 15,
      "id": "grok-4-5-pi",
      "model": "Grok 4.5",
      "harness": "Pi",
      "provider": "SpaceXAI",
      "overallScorePct": 41.3,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 71.01,
        "spatialbench-long": 22.73,
        "scbench": 56.58,
        "scbench-long": 25.76,
        "epibench": 28.93,
        "variantbench": 37.6,
        "txbench-ab": 38.85,
        "txbench-pp": 57.67,
        "txbench-od": 38.35,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 9.3,
        "functionbench": 35.6
      }
    },
    {
      "rank": 16,
      "id": "gpt-5-5-pi",
      "model": "GPT-5.5",
      "harness": "Pi",
      "provider": "OpenAI",
      "overallScorePct": 40.1,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 65.22,
        "spatialbench-long": 30.3,
        "scbench": 56.68,
        "scbench-long": 22.73,
        "epibench": 44.97,
        "variantbench": 29.7,
        "txbench-ab": 33.56,
        "txbench-pp": 55.33,
        "txbench-od": 33.63,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 37.1,
        "functionbench": 29.1
      }
    },
    {
      "rank": 17,
      "id": "claude-opus-4-7-claude-code",
      "model": "Claude Opus 4.7",
      "harness": "Claude Code",
      "provider": "Anthropic",
      "overallScorePct": 39.8,
      "observedBenchmarkCount": 7,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 64.06,
        "spatialbench-long": null,
        "scbench": 54.02,
        "scbench-long": 10.61,
        "epibench": 31.45,
        "variantbench": 35.6,
        "txbench-ab": 39.03,
        "txbench-pp": 44,
        "txbench-od": null,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 33.9,
        "functionbench": null
      }
    },
    {
      "rank": 18,
      "id": "claude-sonnet-5-pi",
      "model": "Claude Sonnet 5",
      "harness": "Pi",
      "provider": "Anthropic",
      "overallScorePct": 39.5,
      "observedBenchmarkCount": 9,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 69.28,
        "spatialbench-long": 24.24,
        "scbench": 59.52,
        "scbench-long": 19.7,
        "epibench": 35.22,
        "variantbench": 38.1,
        "txbench-ab": 33.86,
        "txbench-pp": 42,
        "txbench-od": null,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 30.1,
        "functionbench": 33.8
      }
    },
    {
      "rank": 19,
      "id": "claude-sonnet-5-claude-code",
      "model": "Claude Sonnet 5",
      "harness": "Claude Code",
      "provider": "Anthropic",
      "overallScorePct": 39.1,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 64.35,
        "spatialbench-long": 15.15,
        "scbench": 56.7,
        "scbench-long": 16.67,
        "epibench": 37.74,
        "variantbench": 38.7,
        "txbench-ab": 37.26,
        "txbench-pp": 48.33,
        "txbench-od": 39.29,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 27.4,
        "functionbench": 37.2
      }
    },
    {
      "rank": 20,
      "id": "gpt-5-5-codex",
      "model": "GPT-5.5",
      "harness": "Codex",
      "provider": "OpenAI",
      "overallScorePct": 38.7,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 64.93,
        "spatialbench-long": 33.33,
        "scbench": 57.78,
        "scbench-long": 16.67,
        "epibench": 39.94,
        "variantbench": 28.8,
        "txbench-ab": 31.53,
        "txbench-pp": 47.33,
        "txbench-od": 30.38,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 26.7,
        "functionbench": 36.5
      }
    },
    {
      "rank": 21,
      "id": "kimi-k3-pi",
      "model": "Kimi K3",
      "harness": "Pi",
      "provider": "Moonshot",
      "overallScorePct": 38.1,
      "observedBenchmarkCount": 7,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 61.45,
        "spatialbench-long": 28.79,
        "scbench": 55.73,
        "scbench-long": 16.67,
        "epibench": 26.73,
        "variantbench": 32.8,
        "txbench-ab": null,
        "txbench-pp": 44.67,
        "txbench-od": null,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 8.6,
        "functionbench": null
      }
    },
    {
      "rank": 22,
      "id": "gpt-5-6-luna-pi",
      "model": "GPT-5.6 Luna",
      "harness": "Pi",
      "provider": "OpenAI",
      "overallScorePct": 33.8,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 57.1,
        "spatialbench-long": 21.21,
        "scbench": 51.9,
        "scbench-long": 28.79,
        "epibench": 36.48,
        "variantbench": 28.2,
        "txbench-ab": 19.39,
        "txbench-pp": 46.67,
        "txbench-od": 23.3,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 31.5,
        "functionbench": 25.4
      }
    },
    {
      "rank": 23,
      "id": "gpt-5-6-luna-codex",
      "model": "GPT-5.6 Luna",
      "harness": "Codex",
      "provider": "OpenAI",
      "overallScorePct": 32.2,
      "observedBenchmarkCount": 10,
      "eligibleBenchmarkCount": 11,
      "scores": {
        "spatialbench": 60.58,
        "spatialbench-long": 27.27,
        "scbench": 49.23,
        "scbench-long": 12.12,
        "epibench": 31.45,
        "variantbench": 29.9,
        "txbench-ab": 18.24,
        "txbench-pp": 42,
        "txbench-od": 23.89,
        "biosecbench-surveillance": null,
        "biosecbench-refusal": 33.3,
        "functionbench": 26.9
      }
    }
  ]
}