{
  "generated_at": "2026-03-05T20:14:26.181245+00:00",
  "models": [
    {
      "model_id": "inception/mercury-2",
      "model_slug": "inception-mercury-2",
      "model_label": "inception/mercury-2",
      "overall_score": 64,
      "overall_mark": "F",
      "avg_output_tokens": 830.87,
      "avg_dom_elements": 17.87,
      "run": {
        "run_started_at": "2026-03-05T20:52:34.690656",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 0.35,
          "avg_warning_count": 3.64,
          "avg_output_tokens": 645.29,
          "avg_dom_elements": 15.69,
          "zero_error_count": 122,
          "relative_error_burden_pct": 46.85,
          "relative_warning_burden_pct": 99.27,
          "share_of_model_errors_pct": 19.48,
          "error_share_within_issues_pct": 8.7,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 81.33,
          "score": 64,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "A",
          "mark_distribution_pct": {
            "A": 81.33,
            "B": 16.67,
            "C": 0.67,
            "D": 0.67,
            "E": 0.0,
            "F": 0.67
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 0.69,
          "avg_warning_count": 2.86,
          "avg_output_tokens": 778.1,
          "avg_dom_elements": 17.51,
          "zero_error_count": 94,
          "relative_error_burden_pct": 93.69,
          "relative_warning_burden_pct": 78.0,
          "share_of_model_errors_pct": 38.95,
          "error_share_within_issues_pct": 19.51,
          "improvement_vs_unguided_pct": -100.0,
          "zero_error_share_pct": 62.67,
          "score": 44,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "A",
          "mark_distribution_pct": {
            "A": 62.67,
            "B": 31.33,
            "C": 4.0,
            "D": 1.33,
            "E": 0.67,
            "F": 0.0
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 0.74,
          "avg_warning_count": 3.67,
          "avg_output_tokens": 1063.41,
          "avg_dom_elements": 20.41,
          "zero_error_count": 94,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 41.57,
          "error_share_within_issues_pct": 16.79,
          "improvement_vs_unguided_pct": -113.46,
          "zero_error_share_pct": 62.67,
          "score": 46,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "A",
          "mark_distribution_pct": {
            "A": 62.67,
            "B": 29.33,
            "C": 4.67,
            "D": 2.67,
            "E": 0.67,
            "F": 0.0
          }
        }
      ]
    },
    {
      "model_id": "openai/gpt-5-nano",
      "model_slug": "openai-gpt-5-nano",
      "model_label": "openai/gpt-5-nano",
      "overall_score": 62,
      "overall_mark": "F",
      "avg_output_tokens": 2270.91,
      "avg_dom_elements": 31.2,
      "run": {
        "run_started_at": "2026-02-20T03:40:10.028119",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 1.26,
          "avg_warning_count": 0.96,
          "avg_output_tokens": 2159.17,
          "avg_dom_elements": 28.39,
          "zero_error_count": 81,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 86.23,
          "share_of_model_errors_pct": 49.87,
          "error_share_within_issues_pct": 56.76,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 54.0,
          "score": 38,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "A",
          "mark_distribution_pct": {
            "A": 54.0,
            "B": 36.67,
            "C": 5.33,
            "D": 0.67,
            "E": 0.67,
            "F": 2.67
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 0.43,
          "avg_warning_count": 1.11,
          "avg_output_tokens": 2507.94,
          "avg_dom_elements": 32.94,
          "zero_error_count": 116,
          "relative_error_burden_pct": 34.39,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 17.15,
          "error_share_within_issues_pct": 28.02,
          "improvement_vs_unguided_pct": 65.61,
          "zero_error_share_pct": 77.33,
          "score": 70,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "A",
          "mark_distribution_pct": {
            "A": 77.33,
            "B": 20.0,
            "C": 2.0,
            "D": 0.0,
            "E": 0.0,
            "F": 0.67
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 0.84,
          "avg_warning_count": 0.94,
          "avg_output_tokens": 2143.93,
          "avg_dom_elements": 32.3,
          "zero_error_count": 97,
          "relative_error_burden_pct": 67.03,
          "relative_warning_burden_pct": 84.36,
          "share_of_model_errors_pct": 32.98,
          "error_share_within_issues_pct": 47.35,
          "improvement_vs_unguided_pct": 32.97,
          "zero_error_share_pct": 65.54,
          "score": 53,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "A",
          "mark_distribution_pct": {
            "A": 65.54,
            "B": 25.0,
            "C": 5.41,
            "D": 2.7,
            "E": 0.68,
            "F": 0.68
          }
        }
      ]
    },
    {
      "model_id": "openai/gpt-3.5-turbo",
      "model_slug": "openai-gpt-3-5-turbo",
      "model_label": "openai/gpt-3.5-turbo",
      "overall_score": 49,
      "overall_mark": "F",
      "avg_output_tokens": 355.99,
      "avg_dom_elements": 14.69,
      "run": {
        "run_started_at": "2026-02-20T02:42:49.578237",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 1.4,
          "avg_warning_count": 0.25,
          "avg_output_tokens": 328.71,
          "avg_dom_elements": 13.47,
          "zero_error_count": 75,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 44.4,
          "error_share_within_issues_pct": 85.02,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 50.0,
          "score": 26,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "A",
          "mark_distribution_pct": {
            "A": 50.0,
            "B": 34.0,
            "C": 10.0,
            "D": 1.33,
            "E": 4.67,
            "F": 0.0
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 0.86,
          "avg_warning_count": 0.21,
          "avg_output_tokens": 364.54,
          "avg_dom_elements": 15.29,
          "zero_error_count": 97,
          "relative_error_burden_pct": 61.43,
          "relative_warning_burden_pct": 86.49,
          "share_of_model_errors_pct": 27.27,
          "error_share_within_issues_pct": 80.12,
          "improvement_vs_unguided_pct": 38.57,
          "zero_error_share_pct": 64.67,
          "score": 39,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "A",
          "mark_distribution_pct": {
            "A": 64.67,
            "B": 28.0,
            "C": 5.33,
            "D": 0.0,
            "E": 2.0,
            "F": 0.0
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 0.91,
          "avg_warning_count": 0.2,
          "avg_output_tokens": 374.97,
          "avg_dom_elements": 15.31,
          "zero_error_count": 88,
          "relative_error_burden_pct": 64.67,
          "relative_warning_burden_pct": 79.44,
          "share_of_model_errors_pct": 28.33,
          "error_share_within_issues_pct": 82.21,
          "improvement_vs_unguided_pct": 35.33,
          "zero_error_share_pct": 59.46,
          "score": 36,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "A",
          "mark_distribution_pct": {
            "A": 59.46,
            "B": 32.43,
            "C": 5.41,
            "D": 0.0,
            "E": 2.7,
            "F": 0.0
          }
        }
      ]
    },
    {
      "model_id": "openai/gpt-5.1-codex-mini",
      "model_slug": "openai-gpt-5-1-codex-mini",
      "model_label": "openai/gpt-5.1-codex-mini",
      "overall_score": 46,
      "overall_mark": "F",
      "avg_output_tokens": 1273.82,
      "avg_dom_elements": 23.87,
      "run": {
        "run_started_at": "2026-02-19T22:57:32.235557",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 2.05,
          "avg_warning_count": 0.55,
          "avg_output_tokens": 1124.73,
          "avg_dom_elements": 20.95,
          "zero_error_count": 51,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 90.22,
          "share_of_model_errors_pct": 52.66,
          "error_share_within_issues_pct": 78.72,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 34.0,
          "score": 19,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "A",
          "mark_distribution_pct": {
            "A": 34.0,
            "B": 34.0,
            "C": 20.0,
            "D": 6.67,
            "E": 5.33,
            "F": 0.0
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 0.93,
          "avg_warning_count": 0.61,
          "avg_output_tokens": 1402.59,
          "avg_dom_elements": 24.64,
          "zero_error_count": 87,
          "relative_error_burden_pct": 45.28,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 23.84,
          "error_share_within_issues_pct": 60.17,
          "improvement_vs_unguided_pct": 54.72,
          "zero_error_share_pct": 58.0,
          "score": 43,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "A",
          "mark_distribution_pct": {
            "A": 58.0,
            "B": 32.67,
            "C": 7.33,
            "D": 1.33,
            "E": 0.0,
            "F": 0.67
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 0.93,
          "avg_warning_count": 0.61,
          "avg_output_tokens": 1294.54,
          "avg_dom_elements": 26.05,
          "zero_error_count": 88,
          "relative_error_burden_pct": 45.54,
          "relative_warning_burden_pct": 98.71,
          "share_of_model_errors_pct": 23.5,
          "error_share_within_issues_pct": 60.62,
          "improvement_vs_unguided_pct": 54.46,
          "zero_error_share_pct": 59.86,
          "score": 44,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "A",
          "mark_distribution_pct": {
            "A": 59.86,
            "B": 31.29,
            "C": 6.12,
            "D": 2.04,
            "E": 0.68,
            "F": 0.0
          }
        }
      ]
    },
    {
      "model_id": "openai/gpt-5.3-codex",
      "model_slug": "openai-gpt-5-3-codex",
      "model_label": "openai/gpt-5.3-codex",
      "overall_score": 38,
      "overall_mark": "F",
      "avg_output_tokens": 1631.58,
      "avg_dom_elements": 24.75,
      "run": {
        "run_started_at": "2026-02-25T21:35:20.239366",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 1.93,
          "avg_warning_count": 4.15,
          "avg_output_tokens": 1612.89,
          "avg_dom_elements": 21.69,
          "zero_error_count": 35,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 40.28,
          "error_share_within_issues_pct": 31.8,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 23.33,
          "score": 14,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 23.33,
            "B": 60.67,
            "C": 12.0,
            "D": 1.33,
            "E": 2.0,
            "F": 0.67
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 1.48,
          "avg_warning_count": 2.21,
          "avg_output_tokens": 1733.11,
          "avg_dom_elements": 26.45,
          "zero_error_count": 55,
          "relative_error_burden_pct": 76.55,
          "relative_warning_burden_pct": 53.38,
          "share_of_model_errors_pct": 30.83,
          "error_share_within_issues_pct": 40.07,
          "improvement_vs_unguided_pct": 23.45,
          "zero_error_share_pct": 36.67,
          "score": 26,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 36.67,
            "B": 50.0,
            "C": 10.0,
            "D": 2.0,
            "E": 0.67,
            "F": 0.67
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 1.39,
          "avg_warning_count": 3.42,
          "avg_output_tokens": 1548.73,
          "avg_dom_elements": 26.11,
          "zero_error_count": 47,
          "relative_error_burden_pct": 71.72,
          "relative_warning_burden_pct": 82.48,
          "share_of_model_errors_pct": 28.89,
          "error_share_within_issues_pct": 28.85,
          "improvement_vs_unguided_pct": 28.28,
          "zero_error_share_pct": 31.33,
          "score": 24,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 31.33,
            "B": 55.33,
            "C": 12.67,
            "D": 0.67,
            "E": 0.0,
            "F": 0.0
          }
        }
      ]
    },
    {
      "model_id": "openai/gpt-oss-20b",
      "model_slug": "openai-gpt-oss-20b",
      "model_label": "openai/gpt-oss-20b",
      "overall_score": 28,
      "overall_mark": "F",
      "avg_output_tokens": 616.26,
      "avg_dom_elements": 17.4,
      "run": {
        "run_started_at": "2026-02-20T02:20:47.395548",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 2.3,
          "avg_warning_count": 0.35,
          "avg_output_tokens": 499.3,
          "avg_dom_elements": 15.82,
          "zero_error_count": 10,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 81.71,
          "share_of_model_errors_pct": 39.25,
          "error_share_within_issues_pct": 86.68,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 6.67,
          "score": 4,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 6.67,
            "B": 51.33,
            "C": 33.33,
            "D": 4.67,
            "E": 3.33,
            "F": 0.67
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 1.88,
          "avg_warning_count": 0.36,
          "avg_output_tokens": 652.15,
          "avg_dom_elements": 17.22,
          "zero_error_count": 17,
          "relative_error_burden_pct": 81.74,
          "relative_warning_burden_pct": 83.25,
          "share_of_model_errors_pct": 32.08,
          "error_share_within_issues_pct": 83.93,
          "improvement_vs_unguided_pct": 18.26,
          "zero_error_share_pct": 11.33,
          "score": 7,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 11.33,
            "B": 60.0,
            "C": 22.67,
            "D": 2.67,
            "E": 2.0,
            "F": 1.33
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 1.7,
          "avg_warning_count": 0.43,
          "avg_output_tokens": 698.43,
          "avg_dom_elements": 19.18,
          "zero_error_count": 28,
          "relative_error_burden_pct": 74.03,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 28.67,
          "error_share_within_issues_pct": 79.75,
          "improvement_vs_unguided_pct": 25.97,
          "zero_error_share_pct": 18.92,
          "score": 12,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 18.92,
            "B": 56.76,
            "C": 19.59,
            "D": 2.03,
            "E": 2.03,
            "F": 0.68
          }
        }
      ]
    },
    {
      "model_id": "openai/gpt-4o-2024-11-20",
      "model_slug": "openai-gpt-4o-2024-11-20",
      "model_label": "openai/gpt-4o-2024-11-20",
      "overall_score": 25,
      "overall_mark": "F",
      "avg_output_tokens": 616.13,
      "avg_dom_elements": 16.42,
      "run": {
        "run_started_at": "2026-02-20T03:11:50.598128",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 2.48,
          "avg_warning_count": 0.35,
          "avg_output_tokens": 529.45,
          "avg_dom_elements": 14.97,
          "zero_error_count": 0,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 41.15,
          "error_share_within_issues_pct": 87.74,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.0,
          "score": 0,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 62.0,
            "C": 29.33,
            "D": 4.67,
            "E": 2.67,
            "F": 1.33
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 1.73,
          "avg_warning_count": 0.35,
          "avg_output_tokens": 681.17,
          "avg_dom_elements": 16.41,
          "zero_error_count": 0,
          "relative_error_burden_pct": 69.62,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 28.65,
          "error_share_within_issues_pct": 83.28,
          "improvement_vs_unguided_pct": 30.38,
          "zero_error_share_pct": 0.0,
          "score": 1,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 80.0,
            "C": 16.0,
            "D": 2.67,
            "E": 1.33,
            "F": 0.0
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 1.84,
          "avg_warning_count": 0.24,
          "avg_output_tokens": 638.06,
          "avg_dom_elements": 17.89,
          "zero_error_count": 0,
          "relative_error_burden_pct": 74.38,
          "relative_warning_burden_pct": 68.22,
          "share_of_model_errors_pct": 30.2,
          "error_share_within_issues_pct": 88.64,
          "improvement_vs_unguided_pct": 25.62,
          "zero_error_share_pct": 0.0,
          "score": 1,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 77.7,
            "C": 16.22,
            "D": 2.7,
            "E": 3.38,
            "F": 0.0
          }
        }
      ]
    },
    {
      "model_id": "openai/gpt-5.2",
      "model_slug": "openai-gpt-5-2",
      "model_label": "openai/gpt-5.2",
      "overall_score": 25,
      "overall_mark": "F",
      "avg_output_tokens": 3789.77,
      "avg_dom_elements": 48.43,
      "run": {
        "run_started_at": "2026-02-19T21:23:42.405101",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 3.48,
          "avg_warning_count": 2.26,
          "avg_output_tokens": 4414.03,
          "avg_dom_elements": 46.06,
          "zero_error_count": 0,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 44.27,
          "error_share_within_issues_pct": 60.58,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.0,
          "score": 4,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 52.35,
            "C": 29.53,
            "D": 9.4,
            "E": 4.03,
            "F": 4.7
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 1.98,
          "avg_warning_count": 1.53,
          "avg_output_tokens": 3755.67,
          "avg_dom_elements": 45.85,
          "zero_error_count": 0,
          "relative_error_burden_pct": 56.95,
          "relative_warning_burden_pct": 67.5,
          "share_of_model_errors_pct": 25.38,
          "error_share_within_issues_pct": 56.46,
          "improvement_vs_unguided_pct": 43.05,
          "zero_error_share_pct": 0.0,
          "score": 11,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 74.67,
            "C": 19.33,
            "D": 1.33,
            "E": 3.33,
            "F": 1.33
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 2.4,
          "avg_warning_count": 2.05,
          "avg_output_tokens": 3195.84,
          "avg_dom_elements": 53.41,
          "zero_error_count": 0,
          "relative_error_burden_pct": 69.0,
          "relative_warning_burden_pct": 90.52,
          "share_of_model_errors_pct": 30.34,
          "error_share_within_issues_pct": 53.95,
          "improvement_vs_unguided_pct": 31.0,
          "zero_error_share_pct": 0.0,
          "score": 11,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 74.32,
            "C": 16.89,
            "D": 2.7,
            "E": 2.7,
            "F": 3.38
          }
        }
      ]
    },
    {
      "model_id": "openai/gpt-oss-120b",
      "model_slug": "openai-gpt-oss-120b",
      "model_label": "openai/gpt-oss-120b",
      "overall_score": 25,
      "overall_mark": "F",
      "avg_output_tokens": 850.92,
      "avg_dom_elements": 19.54,
      "run": {
        "run_started_at": "2026-02-20T02:18:52.415818",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 2.72,
          "avg_warning_count": 0.45,
          "avg_output_tokens": 740.85,
          "avg_dom_elements": 17.06,
          "zero_error_count": 0,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 42.72,
          "error_share_within_issues_pct": 85.89,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.0,
          "score": 0,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 54.67,
            "C": 33.33,
            "D": 6.67,
            "E": 2.0,
            "F": 3.33
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 1.83,
          "avg_warning_count": 0.4,
          "avg_output_tokens": 947.69,
          "avg_dom_elements": 20.09,
          "zero_error_count": 3,
          "relative_error_burden_pct": 67.16,
          "relative_warning_burden_pct": 89.55,
          "share_of_model_errors_pct": 28.69,
          "error_share_within_issues_pct": 82.04,
          "improvement_vs_unguided_pct": 32.84,
          "zero_error_share_pct": 2.0,
          "score": 3,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 2.0,
            "B": 66.0,
            "C": 28.0,
            "D": 1.33,
            "E": 2.67,
            "F": 0.0
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 1.84,
          "avg_warning_count": 0.36,
          "avg_output_tokens": 864.39,
          "avg_dom_elements": 21.49,
          "zero_error_count": 1,
          "relative_error_burden_pct": 67.82,
          "relative_warning_burden_pct": 80.17,
          "share_of_model_errors_pct": 28.59,
          "error_share_within_issues_pct": 83.74,
          "improvement_vs_unguided_pct": 32.18,
          "zero_error_share_pct": 0.68,
          "score": 3,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.68,
            "B": 66.22,
            "C": 29.05,
            "D": 0.68,
            "E": 3.38,
            "F": 0.0
          }
        }
      ]
    },
    {
      "model_id": "xiaomi/mimo-v2-flash",
      "model_slug": "xiaomi-mimo-v2-flash",
      "model_label": "xiaomi/mimo-v2-flash",
      "overall_score": 25,
      "overall_mark": "F",
      "avg_output_tokens": 2260.66,
      "avg_dom_elements": 32.05,
      "run": {
        "run_started_at": "2026-02-20T03:35:34.542456",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 3.27,
          "avg_warning_count": 0.5,
          "avg_output_tokens": 1988.14,
          "avg_dom_elements": 26.3,
          "zero_error_count": 2,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 58.14,
          "share_of_model_errors_pct": 46.36,
          "error_share_within_issues_pct": 86.75,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 1.33,
          "score": 1,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 1.33,
            "B": 50.0,
            "C": 36.0,
            "D": 8.67,
            "E": 3.33,
            "F": 0.67
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 1.82,
          "avg_warning_count": 0.86,
          "avg_output_tokens": 2560.47,
          "avg_dom_elements": 34.39,
          "zero_error_count": 1,
          "relative_error_burden_pct": 55.6,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 25.78,
          "error_share_within_issues_pct": 67.91,
          "improvement_vs_unguided_pct": 44.4,
          "zero_error_share_pct": 0.67,
          "score": 8,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.67,
            "B": 71.33,
            "C": 23.33,
            "D": 3.33,
            "E": 1.33,
            "F": 0.0
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 1.99,
          "avg_warning_count": 0.4,
          "avg_output_tokens": 2233.0,
          "avg_dom_elements": 35.5,
          "zero_error_count": 7,
          "relative_error_burden_pct": 60.89,
          "relative_warning_burden_pct": 46.35,
          "share_of_model_errors_pct": 27.86,
          "error_share_within_issues_pct": 83.33,
          "improvement_vs_unguided_pct": 39.11,
          "zero_error_share_pct": 4.73,
          "score": 10,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 4.73,
            "B": 60.81,
            "C": 28.38,
            "D": 3.38,
            "E": 2.7,
            "F": 0.0
          }
        }
      ]
    },
    {
      "model_id": "deepseek/deepseek-v3.2",
      "model_slug": "deepseek-deepseek-v3-2",
      "model_label": "deepseek/deepseek-v3.2",
      "overall_score": 24,
      "overall_mark": "F",
      "avg_output_tokens": 2701.26,
      "avg_dom_elements": 37.03,
      "run": {
        "run_started_at": "2026-02-20T02:33:19.338910",
        "run_status": "cancelled"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 3.4,
          "avg_warning_count": 1.72,
          "avg_output_tokens": 2671.97,
          "avg_dom_elements": 35.35,
          "zero_error_count": 0,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 46.58,
          "error_share_within_issues_pct": 66.41,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.0,
          "score": 2,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 41.33,
            "C": 37.33,
            "D": 10.67,
            "E": 8.0,
            "F": 2.67
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 1.97,
          "avg_warning_count": 1.0,
          "avg_output_tokens": 2992.02,
          "avg_dom_elements": 36.09,
          "zero_error_count": 0,
          "relative_error_burden_pct": 57.84,
          "relative_warning_burden_pct": 58.14,
          "share_of_model_errors_pct": 26.76,
          "error_share_within_issues_pct": 66.29,
          "improvement_vs_unguided_pct": 42.16,
          "zero_error_share_pct": 0.0,
          "score": 8,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 75.84,
            "C": 18.12,
            "D": 3.36,
            "E": 2.01,
            "F": 0.67
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 1.99,
          "avg_warning_count": 0.8,
          "avg_output_tokens": 2436.43,
          "avg_dom_elements": 39.69,
          "zero_error_count": 0,
          "relative_error_burden_pct": 58.42,
          "relative_warning_burden_pct": 46.27,
          "share_of_model_errors_pct": 26.67,
          "error_share_within_issues_pct": 71.39,
          "improvement_vs_unguided_pct": 41.58,
          "zero_error_share_pct": 0.0,
          "score": 9,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 74.83,
            "C": 19.73,
            "D": 2.04,
            "E": 2.72,
            "F": 0.68
          }
        }
      ]
    },
    {
      "model_id": "google/gemma-3-12b-it",
      "model_slug": "google-gemma-3-12b-it",
      "model_label": "google/gemma-3-12b-it",
      "overall_score": 24,
      "overall_mark": "F",
      "avg_output_tokens": 813.11,
      "avg_dom_elements": 18.92,
      "run": {
        "run_started_at": "2026-02-20T03:02:21.339330",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 2.54,
          "avg_warning_count": 0.21,
          "avg_output_tokens": 697.49,
          "avg_dom_elements": 16.6,
          "zero_error_count": 4,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 57.41,
          "share_of_model_errors_pct": 37.5,
          "error_share_within_issues_pct": 92.25,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 2.67,
          "score": 2,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 2.67,
            "B": 58.0,
            "C": 33.33,
            "D": 5.33,
            "E": 0.0,
            "F": 0.67
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 1.81,
          "avg_warning_count": 0.35,
          "avg_output_tokens": 881.01,
          "avg_dom_elements": 19.22,
          "zero_error_count": 3,
          "relative_error_burden_pct": 71.13,
          "relative_warning_burden_pct": 93.28,
          "share_of_model_errors_pct": 26.67,
          "error_share_within_issues_pct": 83.9,
          "improvement_vs_unguided_pct": 28.87,
          "zero_error_share_pct": 2.0,
          "score": 3,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 2.0,
            "B": 67.33,
            "C": 28.0,
            "D": 1.33,
            "E": 0.67,
            "F": 0.67
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 2.46,
          "avg_warning_count": 0.37,
          "avg_output_tokens": 861.47,
          "avg_dom_elements": 20.98,
          "zero_error_count": 0,
          "relative_error_burden_pct": 96.83,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 35.83,
          "error_share_within_issues_pct": 86.87,
          "improvement_vs_unguided_pct": 3.17,
          "zero_error_share_pct": 0.0,
          "score": 1,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 63.51,
            "C": 27.7,
            "D": 4.05,
            "E": 2.7,
            "F": 2.03
          }
        }
      ]
    },
    {
      "model_id": "anthropic/claude-3-haiku",
      "model_slug": "anthropic-claude-3-haiku",
      "model_label": "anthropic/claude-3-haiku",
      "overall_score": 23,
      "overall_mark": "F",
      "avg_output_tokens": 883.4,
      "avg_dom_elements": 17.04,
      "run": {
        "run_started_at": "2026-02-20T02:53:30.311116",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 2.45,
          "avg_warning_count": 0.3,
          "avg_output_tokens": 745.91,
          "avg_dom_elements": 14.96,
          "zero_error_count": 0,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 70.48,
          "share_of_model_errors_pct": 36.77,
          "error_share_within_issues_pct": 89.08,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.0,
          "score": 0,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 56.0,
            "C": 38.0,
            "D": 3.33,
            "E": 1.33,
            "F": 1.33
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 2.03,
          "avg_warning_count": 0.24,
          "avg_output_tokens": 944.85,
          "avg_dom_elements": 17.53,
          "zero_error_count": 0,
          "relative_error_burden_pct": 83.12,
          "relative_warning_burden_pct": 56.76,
          "share_of_model_errors_pct": 30.36,
          "error_share_within_issues_pct": 89.38,
          "improvement_vs_unguided_pct": 16.88,
          "zero_error_share_pct": 0.0,
          "score": 1,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 67.79,
            "C": 28.86,
            "D": 1.34,
            "E": 1.34,
            "F": 0.67
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 2.22,
          "avg_warning_count": 0.43,
          "avg_output_tokens": 960.87,
          "avg_dom_elements": 18.64,
          "zero_error_count": 0,
          "relative_error_burden_pct": 90.58,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 32.87,
          "error_share_within_issues_pct": 83.89,
          "improvement_vs_unguided_pct": 9.42,
          "zero_error_share_pct": 0.0,
          "score": 1,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 67.57,
            "C": 25.68,
            "D": 1.35,
            "E": 4.73,
            "F": 0.68
          }
        }
      ]
    },
    {
      "model_id": "google/gemini-3-flash-preview",
      "model_slug": "google-gemini-3-flash-preview",
      "model_label": "google/gemini-3-flash-preview",
      "overall_score": 23,
      "overall_mark": "F",
      "avg_output_tokens": 1168.64,
      "avg_dom_elements": 20.36,
      "run": {
        "run_started_at": "2026-02-20T01:16:52.379305",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 3.43,
          "avg_warning_count": 0.53,
          "avg_output_tokens": 993.9,
          "avg_dom_elements": 17.59,
          "zero_error_count": 0,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 49.57,
          "error_share_within_issues_pct": 86.7,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.0,
          "score": 0,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 45.33,
            "C": 34.0,
            "D": 12.67,
            "E": 4.67,
            "F": 3.33
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 1.73,
          "avg_warning_count": 0.34,
          "avg_output_tokens": 1356.41,
          "avg_dom_elements": 21.46,
          "zero_error_count": 0,
          "relative_error_burden_pct": 50.29,
          "relative_warning_burden_pct": 64.56,
          "share_of_model_errors_pct": 24.93,
          "error_share_within_issues_pct": 83.55,
          "improvement_vs_unguided_pct": 49.71,
          "zero_error_share_pct": 0.0,
          "score": 3,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 69.33,
            "C": 28.67,
            "D": 2.0,
            "E": 0.0,
            "F": 0.0
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 1.79,
          "avg_warning_count": 0.45,
          "avg_output_tokens": 1155.43,
          "avg_dom_elements": 22.07,
          "zero_error_count": 0,
          "relative_error_burden_pct": 52.15,
          "relative_warning_burden_pct": 85.96,
          "share_of_model_errors_pct": 25.51,
          "error_share_within_issues_pct": 79.82,
          "improvement_vs_unguided_pct": 47.85,
          "zero_error_share_pct": 0.0,
          "score": 3,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 71.62,
            "C": 22.97,
            "D": 4.05,
            "E": 1.35,
            "F": 0.0
          }
        }
      ]
    },
    {
      "model_id": "qwen/qwen3-coder-next",
      "model_slug": "qwen-qwen3-coder-next",
      "model_label": "qwen/qwen3-coder-next",
      "overall_score": 23,
      "overall_mark": "F",
      "avg_output_tokens": 2665.93,
      "avg_dom_elements": 38.35,
      "run": {
        "run_started_at": "2026-02-20T01:08:09.141791",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 3.43,
          "avg_warning_count": 0.86,
          "avg_output_tokens": 2486.57,
          "avg_dom_elements": 34.73,
          "zero_error_count": 1,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 44.2,
          "error_share_within_issues_pct": 79.94,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.67,
          "score": 2,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.67,
            "B": 51.33,
            "C": 37.33,
            "D": 5.33,
            "E": 4.0,
            "F": 1.33
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 2.17,
          "avg_warning_count": 0.83,
          "avg_output_tokens": 3031.37,
          "avg_dom_elements": 40.28,
          "zero_error_count": 0,
          "relative_error_burden_pct": 63.42,
          "relative_warning_burden_pct": 96.12,
          "share_of_model_errors_pct": 28.03,
          "error_share_within_issues_pct": 72.44,
          "improvement_vs_unguided_pct": 36.58,
          "zero_error_share_pct": 0.0,
          "score": 8,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 71.33,
            "C": 22.0,
            "D": 2.67,
            "E": 2.67,
            "F": 1.33
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 2.18,
          "avg_warning_count": 0.76,
          "avg_output_tokens": 2477.34,
          "avg_dom_elements": 40.08,
          "zero_error_count": 0,
          "relative_error_burden_pct": 63.69,
          "relative_warning_burden_pct": 88.78,
          "share_of_model_errors_pct": 27.77,
          "error_share_within_issues_pct": 74.08,
          "improvement_vs_unguided_pct": 36.31,
          "zero_error_share_pct": 0.0,
          "score": 8,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 69.59,
            "C": 19.59,
            "D": 4.73,
            "E": 5.41,
            "F": 0.68
          }
        }
      ]
    },
    {
      "model_id": "x-ai/grok-4.1-fast",
      "model_slug": "x-ai-grok-4-1-fast",
      "model_label": "x-ai/grok-4.1-fast",
      "overall_score": 22,
      "overall_mark": "F",
      "avg_output_tokens": 1552.35,
      "avg_dom_elements": 28.48,
      "run": {
        "run_started_at": "2026-02-20T00:48:52.832567",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 3.18,
          "avg_warning_count": 0.84,
          "avg_output_tokens": 1337.3,
          "avg_dom_elements": 24.95,
          "zero_error_count": 0,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 42.48,
          "error_share_within_issues_pct": 79.1,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.0,
          "score": 1,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 50.67,
            "C": 31.33,
            "D": 8.67,
            "E": 5.33,
            "F": 4.0
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 2.19,
          "avg_warning_count": 0.63,
          "avg_output_tokens": 1838.4,
          "avg_dom_elements": 28.9,
          "zero_error_count": 0,
          "relative_error_burden_pct": 68.97,
          "relative_warning_burden_pct": 75.4,
          "share_of_model_errors_pct": 29.3,
          "error_share_within_issues_pct": 77.59,
          "improvement_vs_unguided_pct": 31.03,
          "zero_error_share_pct": 0.0,
          "score": 4,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 76.67,
            "C": 16.0,
            "D": 2.0,
            "E": 2.67,
            "F": 2.67
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 2.14,
          "avg_warning_count": 0.64,
          "avg_output_tokens": 1480.4,
          "avg_dom_elements": 31.64,
          "zero_error_count": 0,
          "relative_error_burden_pct": 67.36,
          "relative_warning_burden_pct": 76.42,
          "share_of_model_errors_pct": 28.23,
          "error_share_within_issues_pct": 76.94,
          "improvement_vs_unguided_pct": 32.64,
          "zero_error_share_pct": 0.0,
          "score": 5,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 70.95,
            "C": 19.59,
            "D": 4.05,
            "E": 4.73,
            "F": 0.68
          }
        }
      ]
    },
    {
      "model_id": "mistralai/devstral-2512",
      "model_slug": "mistralai-devstral-2512",
      "model_label": "mistralai/devstral-2512",
      "overall_score": 21,
      "overall_mark": "F",
      "avg_output_tokens": 1327.05,
      "avg_dom_elements": 24.61,
      "run": {
        "run_started_at": "2026-02-20T01:10:49.070322",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 3.57,
          "avg_warning_count": 0.68,
          "avg_output_tokens": 1036.63,
          "avg_dom_elements": 20.72,
          "zero_error_count": 0,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 45.57,
          "error_share_within_issues_pct": 83.99,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.0,
          "score": 0,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 46.67,
            "C": 36.67,
            "D": 10.67,
            "E": 3.33,
            "F": 2.67
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 1.99,
          "avg_warning_count": 0.51,
          "avg_output_tokens": 1617.75,
          "avg_dom_elements": 26.86,
          "zero_error_count": 0,
          "relative_error_burden_pct": 55.89,
          "relative_warning_burden_pct": 74.51,
          "share_of_model_errors_pct": 25.47,
          "error_share_within_issues_pct": 79.73,
          "improvement_vs_unguided_pct": 44.11,
          "zero_error_share_pct": 0.0,
          "score": 4,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 66.67,
            "C": 28.0,
            "D": 2.0,
            "E": 2.67,
            "F": 0.67
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 2.3,
          "avg_warning_count": 0.51,
          "avg_output_tokens": 1326.78,
          "avg_dom_elements": 26.26,
          "zero_error_count": 0,
          "relative_error_burden_pct": 64.41,
          "relative_warning_burden_pct": 74.52,
          "share_of_model_errors_pct": 28.96,
          "error_share_within_issues_pct": 81.93,
          "improvement_vs_unguided_pct": 35.59,
          "zero_error_share_pct": 0.0,
          "score": 2,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 62.84,
            "C": 26.35,
            "D": 6.08,
            "E": 4.73,
            "F": 0.0
          }
        }
      ]
    },
    {
      "model_id": "anthropic/claude-haiku-4.5",
      "model_slug": "anthropic-claude-haiku-4-5",
      "model_label": "anthropic/claude-haiku-4.5",
      "overall_score": 20,
      "overall_mark": "F",
      "avg_output_tokens": 2260.72,
      "avg_dom_elements": 33.9,
      "run": {
        "run_started_at": "2026-02-20T03:47:47.346614",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 3.28,
          "avg_warning_count": 0.97,
          "avg_output_tokens": 1905.26,
          "avg_dom_elements": 27.59,
          "zero_error_count": 0,
          "relative_error_burden_pct": 92.36,
          "relative_warning_burden_pct": 81.71,
          "share_of_model_errors_pct": 38.53,
          "error_share_within_issues_pct": 77.12,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.0,
          "score": 1,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 52.67,
            "C": 27.33,
            "D": 12.67,
            "E": 4.0,
            "F": 3.33
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 2.01,
          "avg_warning_count": 0.85,
          "avg_output_tokens": 2243.37,
          "avg_dom_elements": 29.03,
          "zero_error_count": 0,
          "relative_error_burden_pct": 56.69,
          "relative_warning_burden_pct": 71.08,
          "share_of_model_errors_pct": 23.65,
          "error_share_within_issues_pct": 70.4,
          "improvement_vs_unguided_pct": 38.62,
          "zero_error_share_pct": 0.0,
          "score": 5,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 76.67,
            "C": 14.67,
            "D": 4.67,
            "E": 2.0,
            "F": 2.0
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 3.55,
          "avg_warning_count": 1.19,
          "avg_output_tokens": 2671.9,
          "avg_dom_elements": 46.22,
          "zero_error_count": 0,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 37.82,
          "error_share_within_issues_pct": 74.88,
          "improvement_vs_unguided_pct": -8.28,
          "zero_error_share_pct": 0.0,
          "score": 3,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 70.59,
            "C": 14.71,
            "D": 8.09,
            "E": 2.21,
            "F": 4.41
          }
        }
      ]
    },
    {
      "model_id": "anthropic/claude-opus-4.6",
      "model_slug": "anthropic-claude-opus-4-6",
      "model_label": "anthropic/claude-opus-4.6",
      "overall_score": 20,
      "overall_mark": "F",
      "avg_output_tokens": 4617.76,
      "avg_dom_elements": 56.81,
      "run": {
        "run_started_at": "2026-02-20T01:56:59.037152",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 5.17,
          "avg_warning_count": 1.78,
          "avg_output_tokens": 3900.16,
          "avg_dom_elements": 43.79,
          "zero_error_count": 0,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 52.15,
          "error_share_within_issues_pct": 74.4,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.0,
          "score": 1,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 43.33,
            "C": 26.0,
            "D": 13.33,
            "E": 9.33,
            "F": 8.0
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 2.02,
          "avg_warning_count": 0.99,
          "avg_output_tokens": 4497.89,
          "avg_dom_elements": 48.4,
          "zero_error_count": 0,
          "relative_error_burden_pct": 39.05,
          "relative_warning_burden_pct": 55.43,
          "share_of_model_errors_pct": 20.36,
          "error_share_within_issues_pct": 67.18,
          "improvement_vs_unguided_pct": 60.95,
          "zero_error_share_pct": 0.0,
          "score": 12,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 69.33,
            "C": 26.0,
            "D": 0.67,
            "E": 3.33,
            "F": 0.67
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 2.76,
          "avg_warning_count": 1.72,
          "avg_output_tokens": 5466.55,
          "avg_dom_elements": 78.55,
          "zero_error_count": 0,
          "relative_error_burden_pct": 53.42,
          "relative_warning_burden_pct": 96.8,
          "share_of_model_errors_pct": 27.49,
          "error_share_within_issues_pct": 61.6,
          "improvement_vs_unguided_pct": 46.58,
          "zero_error_share_pct": 0.0,
          "score": 15,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 70.27,
            "C": 16.89,
            "D": 2.7,
            "E": 5.41,
            "F": 4.73
          }
        }
      ]
    },
    {
      "model_id": "google/gemini-3.1-pro-preview",
      "model_slug": "google-gemini-3-1-pro-preview",
      "model_label": "google/gemini-3.1-pro-preview",
      "overall_score": 18,
      "overall_mark": "F",
      "avg_output_tokens": 1193.78,
      "avg_dom_elements": 15.79,
      "run": {
        "run_started_at": "2026-02-25T21:13:06.949976",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 3.52,
          "avg_warning_count": 0.25,
          "avg_output_tokens": 1084.17,
          "avg_dom_elements": 14.57,
          "zero_error_count": 1,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 75.51,
          "share_of_model_errors_pct": 39.43,
          "error_share_within_issues_pct": 93.45,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.67,
          "score": 0,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "C",
          "mark_distribution_pct": {
            "A": 0.67,
            "B": 38.0,
            "C": 47.33,
            "D": 8.67,
            "E": 4.0,
            "F": 1.33
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 2.78,
          "avg_warning_count": 0.33,
          "avg_output_tokens": 1348.94,
          "avg_dom_elements": 15.86,
          "zero_error_count": 0,
          "relative_error_burden_pct": 78.98,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 31.14,
          "error_share_within_issues_pct": 89.48,
          "improvement_vs_unguided_pct": 21.02,
          "zero_error_share_pct": 0.0,
          "score": 0,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 61.33,
            "C": 35.33,
            "D": 1.33,
            "E": 2.0,
            "F": 0.0
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 2.68,
          "avg_warning_count": 0.31,
          "avg_output_tokens": 1147.3,
          "avg_dom_elements": 16.96,
          "zero_error_count": 1,
          "relative_error_burden_pct": 76.14,
          "relative_warning_burden_pct": 93.71,
          "share_of_model_errors_pct": 29.42,
          "error_share_within_issues_pct": 89.75,
          "improvement_vs_unguided_pct": 23.86,
          "zero_error_share_pct": 0.68,
          "score": 1,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 0.68,
            "B": 64.63,
            "C": 27.89,
            "D": 5.44,
            "E": 1.36,
            "F": 0.0
          }
        }
      ]
    },
    {
      "model_id": "z-ai/glm-5",
      "model_slug": "z-ai-glm-5",
      "model_label": "z-ai/glm-5",
      "overall_score": 14,
      "overall_mark": "F",
      "avg_output_tokens": 4314.77,
      "avg_dom_elements": 59.35,
      "run": {
        "run_started_at": "2026-02-20T01:15:08.811793",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 4.75,
          "avg_warning_count": 1.75,
          "avg_output_tokens": 4397.91,
          "avg_dom_elements": 54.73,
          "zero_error_count": 1,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 37.61,
          "error_share_within_issues_pct": 73.13,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 0.67,
          "score": 3,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "D",
          "mark_distribution_pct": {
            "A": 0.67,
            "B": 21.33,
            "C": 26.67,
            "D": 35.33,
            "E": 14.0,
            "F": 2.0
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 4.23,
          "avg_warning_count": 1.24,
          "avg_output_tokens": 4302.16,
          "avg_dom_elements": 55.79,
          "zero_error_count": 0,
          "relative_error_burden_pct": 88.98,
          "relative_warning_burden_pct": 71.18,
          "share_of_model_errors_pct": 33.02,
          "error_share_within_issues_pct": 77.28,
          "improvement_vs_unguided_pct": 11.02,
          "zero_error_share_pct": 0.0,
          "score": 4,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "D",
          "mark_distribution_pct": {
            "A": 0.0,
            "B": 25.0,
            "C": 30.41,
            "D": 36.49,
            "E": 7.43,
            "F": 0.68
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 3.79,
          "avg_warning_count": 1.25,
          "avg_output_tokens": 4242.61,
          "avg_dom_elements": 67.65,
          "zero_error_count": 2,
          "relative_error_burden_pct": 79.71,
          "relative_warning_burden_pct": 71.66,
          "share_of_model_errors_pct": 29.38,
          "error_share_within_issues_pct": 75.17,
          "improvement_vs_unguided_pct": 20.29,
          "zero_error_share_pct": 1.36,
          "score": 8,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 1.36,
            "B": 36.05,
            "C": 31.97,
            "D": 23.13,
            "E": 6.12,
            "F": 1.36
          }
        }
      ]
    },
    {
      "model_id": "moonshotai/kimi-k2.5",
      "model_slug": "moonshotai-kimi-k2-5",
      "model_label": "moonshotai/kimi-k2.5",
      "overall_score": 11,
      "overall_mark": "F",
      "avg_output_tokens": 4004.76,
      "avg_dom_elements": 64.69,
      "run": {
        "run_started_at": "2026-02-19T22:21:53.345272",
        "run_status": "completed"
      },
      "guidance": [
        {
          "label": "Unguided",
          "avg_error_count": 7.11,
          "avg_warning_count": 4.46,
          "avg_output_tokens": 4312.72,
          "avg_dom_elements": 66.21,
          "zero_error_count": 2,
          "relative_error_burden_pct": 100.0,
          "relative_warning_burden_pct": 100.0,
          "share_of_model_errors_pct": 47.05,
          "error_share_within_issues_pct": 61.45,
          "improvement_vs_unguided_pct": 0.0,
          "zero_error_share_pct": 1.34,
          "score": 2,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 1.34,
            "B": 20.81,
            "C": 19.46,
            "D": 17.45,
            "E": 20.81,
            "F": 20.13
          }
        },
        {
          "label": "Little guidance",
          "avg_error_count": 4.53,
          "avg_warning_count": 3.86,
          "avg_output_tokens": 4209.81,
          "avg_dom_elements": 66.23,
          "zero_error_count": 3,
          "relative_error_burden_pct": 63.72,
          "relative_warning_burden_pct": 86.49,
          "share_of_model_errors_pct": 30.18,
          "error_share_within_issues_pct": 54.01,
          "improvement_vs_unguided_pct": 36.28,
          "zero_error_share_pct": 2.0,
          "score": 6,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 2.0,
            "B": 43.33,
            "C": 28.67,
            "D": 8.67,
            "E": 10.0,
            "F": 7.33
          }
        },
        {
          "label": "Expert guidance",
          "avg_error_count": 3.66,
          "avg_warning_count": 1.46,
          "avg_output_tokens": 3457.3,
          "avg_dom_elements": 61.42,
          "zero_error_count": 2,
          "relative_error_burden_pct": 51.51,
          "relative_warning_burden_pct": 32.65,
          "share_of_model_errors_pct": 22.77,
          "error_share_within_issues_pct": 71.55,
          "improvement_vs_unguided_pct": 48.49,
          "zero_error_share_pct": 1.43,
          "score": 7,
          "mark": "F",
          "dominant_mark": "F",
          "dominant_mark_from_distribution": "B",
          "mark_distribution_pct": {
            "A": 1.43,
            "B": 50.0,
            "C": 26.43,
            "D": 7.86,
            "E": 6.43,
            "F": 7.86
          }
        }
      ]
    }
  ]
}
