{
  "artifact_sha256": "2f99cec4e2e79029a81b8a601f76cf5bba5b237809e9d6d5981d7adc9153adb8",
  "combined_inventory": {
    "complete_pairs": 211,
    "completed_response_arms": 524,
    "distinct_tasks": 28,
    "epicure_calls": 488,
    "epicure_successful_calls": 380,
    "known_conservative_exposure_subtotal_usd": 53.707710526666666,
    "maximum_complete_pairs_per_model": 19,
    "minimum_complete_pairs_per_model": 8,
    "models": 16,
    "models_with_at_least_eight_complete_pairs": 16,
    "provider_charge_complete": false,
    "provider_generation_ids": 1883,
    "quality_judgments": 0,
    "scheduled_pairs": 320,
    "synthetic_tasks": 0,
    "unpriced_model_ids": [
      "cohere/command-a-plus-05-2026",
      "cohere/command-a-reasoning-08-2025"
    ]
  },
  "family_rows": [
    {
      "descriptive_difference_percentage_points": 25.563909774436095,
      "display_name": "Substitution",
      "family": "substitution",
      "high_resource_complete_pairs": 26,
      "high_resource_completion_rate": 0.6842105263157895,
      "high_resource_scheduled_pairs": 38,
      "high_resource_wilson_lower_95": 0.5254424010939046,
      "high_resource_wilson_upper_95": 0.8091539632440382,
      "strict_complete_pairs": 18,
      "strict_completion_rate": 0.42857142857142855,
      "strict_scheduled_pairs": 42,
      "strict_wilson_lower_95": 0.2911765117618606,
      "strict_wilson_upper_95": 0.5779376012189541,
      "total_complete_pairs": 44,
      "total_scheduled_pairs": 80
    },
    {
      "descriptive_difference_percentage_points": 5.388471177944865,
      "display_name": "Composition",
      "family": "composition",
      "high_resource_complete_pairs": 31,
      "high_resource_completion_rate": 0.8157894736842105,
      "high_resource_scheduled_pairs": 38,
      "high_resource_wilson_lower_95": 0.665811419288686,
      "high_resource_wilson_upper_95": 0.9077823481477872,
      "strict_complete_pairs": 32,
      "strict_completion_rate": 0.7619047619047619,
      "strict_scheduled_pairs": 42,
      "strict_wilson_lower_95": 0.6147245613168169,
      "strict_wilson_upper_95": 0.8651903577535287,
      "total_complete_pairs": 63,
      "total_scheduled_pairs": 80
    },
    {
      "descriptive_difference_percentage_points": 6.015037593984962,
      "display_name": "Cookability",
      "family": "cookability",
      "high_resource_complete_pairs": 24,
      "high_resource_completion_rate": 0.631578947368421,
      "high_resource_scheduled_pairs": 38,
      "high_resource_wilson_lower_95": 0.47283959557009003,
      "high_resource_wilson_upper_95": 0.7661578075284404,
      "strict_complete_pairs": 24,
      "strict_completion_rate": 0.5714285714285714,
      "strict_scheduled_pairs": 42,
      "strict_wilson_lower_95": 0.4220623987810458,
      "strict_wilson_upper_95": 0.7088234882381393,
      "total_complete_pairs": 48,
      "total_scheduled_pairs": 80
    },
    {
      "descriptive_difference_percentage_points": 12.030075187969924,
      "display_name": "Evidence interpretation",
      "family": "evidence",
      "high_resource_complete_pairs": 29,
      "high_resource_completion_rate": 0.7631578947368421,
      "high_resource_scheduled_pairs": 38,
      "high_resource_wilson_lower_95": 0.607931876367602,
      "high_resource_wilson_upper_95": 0.8700629298294591,
      "strict_complete_pairs": 27,
      "strict_completion_rate": 0.6428571428571429,
      "strict_scheduled_pairs": 42,
      "strict_wilson_lower_95": 0.4916636571945582,
      "strict_wilson_upper_95": 0.7701081168438123,
      "total_complete_pairs": 56,
      "total_scheduled_pairs": 80
    }
  ],
  "high_resource": {
    "complete_pairs": 110,
    "completed_response_arms": 262,
    "distinct_tasks": 16,
    "epicure_calls": 273,
    "epicure_successful_calls": 207,
    "execution_policy_sha256": "579bef8dee7495d1b695c7d59365a218afebedaeb71cbad136eaab9e28d5916d",
    "known_conservative_exposure_subtotal_usd": 27.162050533333332,
    "provider_charge_complete": false,
    "provider_generation_ids": 919,
    "scheduled_pairs": 152,
    "source_artifact_sha256": "c0bd526a2776a25adfbd2c43b98b8f15c143a8cb93b957ba961d0e9efe626688",
    "task_set_sha256": "dc6a5ecea427ac2fec8198be9a852d1ca29d4033cd26bee9f240bd7c19fb2a92",
    "unpriced_model_ids": [
      "cohere/command-a-plus-05-2026",
      "cohere/command-a-reasoning-08-2025"
    ]
  },
  "interpretation": "descriptive_operational_completion_only",
  "limitations": [
    "The task sets are disjoint, so between-stratum differences are not causal resource effects.",
    "Pair completion is an execution-reliability measure, not a model-quality score.",
    "Wilson intervals are descriptive binomial intervals and do not model task clustering.",
    "No preference or rubric judgments are present in either stratum.",
    "The combined USD cost is a known-exposure subtotal because direct Cohere provider charges were not returned.",
    "The two Cohere endpoints were prespecified as a high-resource extension and were not run in the earlier strict stratum."
  ],
  "model_rows": [
    {
      "canonical_model_slug": "openai/gpt-5.6-sol-pro-20260709",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": 16.666666666666664,
      "display_name": "GPT-5.6 Sol (OR pro)",
      "high_resource_complete_pairs": 6,
      "high_resource_completion_rate": 0.75,
      "high_resource_conservative_cost_exposure_usd": 3.129580333333333,
      "high_resource_epicure_calls": 10,
      "high_resource_epicure_successful_calls": 10,
      "high_resource_scheduled_pairs": 8,
      "high_resource_wilson_lower_95": 0.40927543031016883,
      "high_resource_wilson_upper_95": 0.9285207872478909,
      "manifest_ordinal": 1,
      "minimum_eight_complete_pairs": true,
      "model_id": "openai/gpt-5.6-sol-pro",
      "provider_tag": "openai/flex",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 7,
      "strict_completion_rate": 0.5833333333333334,
      "strict_conservative_cost_exposure_usd": 2.279244,
      "strict_epicure_calls": 10,
      "strict_epicure_successful_calls": 9,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.31951131254954973,
      "strict_wilson_upper_95": 0.8067396863412435,
      "total_complete_pairs": 13,
      "total_conservative_cost_exposure_usd": 5.4088243333333335,
      "total_epicure_calls": 20,
      "total_epicure_successful_calls": 19,
      "total_scheduled_pairs": 20
    },
    {
      "canonical_model_slug": "anthropic/claude-5-fable-20260609",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": 4.1666666666666625,
      "display_name": "Claude Fable 5",
      "high_resource_complete_pairs": 7,
      "high_resource_completion_rate": 0.875,
      "high_resource_conservative_cost_exposure_usd": 2.824137,
      "high_resource_epicure_calls": 26,
      "high_resource_epicure_successful_calls": 20,
      "high_resource_scheduled_pairs": 8,
      "high_resource_wilson_lower_95": 0.5291118177871464,
      "high_resource_wilson_upper_95": 0.9775825085499433,
      "manifest_ordinal": 2,
      "minimum_eight_complete_pairs": true,
      "model_id": "anthropic/claude-fable-5",
      "provider_tag": "anthropic",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 10,
      "strict_completion_rate": 0.8333333333333334,
      "strict_conservative_cost_exposure_usd": 3.860119,
      "strict_epicure_calls": 33,
      "strict_epicure_successful_calls": 23,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.5519691377470266,
      "strict_wilson_upper_95": 0.9530348578161463,
      "total_complete_pairs": 17,
      "total_conservative_cost_exposure_usd": 6.6842559999999995,
      "total_epicure_calls": 59,
      "total_epicure_successful_calls": 43,
      "total_scheduled_pairs": 20
    },
    {
      "canonical_model_slug": "anthropic/claude-opus-5-20260723",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": 8.333333333333337,
      "display_name": "Claude Opus 5",
      "high_resource_complete_pairs": 8,
      "high_resource_completion_rate": 1.0,
      "high_resource_conservative_cost_exposure_usd": 2.38987,
      "high_resource_epicure_calls": 16,
      "high_resource_epicure_successful_calls": 13,
      "high_resource_scheduled_pairs": 8,
      "high_resource_wilson_lower_95": 0.6755924351161198,
      "high_resource_wilson_upper_95": 1.0,
      "manifest_ordinal": 3,
      "minimum_eight_complete_pairs": true,
      "model_id": "anthropic/claude-opus-5",
      "provider_tag": "anthropic",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 11,
      "strict_completion_rate": 0.9166666666666666,
      "strict_conservative_cost_exposure_usd": 3.373052,
      "strict_epicure_calls": 23,
      "strict_epicure_successful_calls": 20,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.646120088858883,
      "strict_wilson_upper_95": 0.9851349055950829,
      "total_complete_pairs": 19,
      "total_conservative_cost_exposure_usd": 5.762922,
      "total_epicure_calls": 39,
      "total_epicure_successful_calls": 33,
      "total_scheduled_pairs": 20
    },
    {
      "canonical_model_slug": "anthropic/claude-sonnet-5-20260630",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": -4.1666666666666625,
      "display_name": "Claude Sonnet 5",
      "high_resource_complete_pairs": 7,
      "high_resource_completion_rate": 0.875,
      "high_resource_conservative_cost_exposure_usd": 0.765126,
      "high_resource_epicure_calls": 30,
      "high_resource_epicure_successful_calls": 20,
      "high_resource_scheduled_pairs": 8,
      "high_resource_wilson_lower_95": 0.5291118177871464,
      "high_resource_wilson_upper_95": 0.9775825085499433,
      "manifest_ordinal": 4,
      "minimum_eight_complete_pairs": true,
      "model_id": "anthropic/claude-sonnet-5",
      "provider_tag": "anthropic",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 11,
      "strict_completion_rate": 0.9166666666666666,
      "strict_conservative_cost_exposure_usd": 0.94377,
      "strict_epicure_calls": 21,
      "strict_epicure_successful_calls": 17,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.646120088858883,
      "strict_wilson_upper_95": 0.9851349055950829,
      "total_complete_pairs": 18,
      "total_conservative_cost_exposure_usd": 1.708896,
      "total_epicure_calls": 51,
      "total_epicure_successful_calls": 37,
      "total_scheduled_pairs": 20
    },
    {
      "canonical_model_slug": "google/gemini-3.1-pro-preview-20260219",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": 8.333333333333337,
      "display_name": "Gemini 3.1 Pro",
      "high_resource_complete_pairs": 6,
      "high_resource_completion_rate": 0.75,
      "high_resource_conservative_cost_exposure_usd": 0.30485,
      "high_resource_epicure_calls": 6,
      "high_resource_epicure_successful_calls": 6,
      "high_resource_scheduled_pairs": 8,
      "high_resource_wilson_lower_95": 0.40927543031016883,
      "high_resource_wilson_upper_95": 0.9285207872478909,
      "manifest_ordinal": 5,
      "minimum_eight_complete_pairs": true,
      "model_id": "google/gemini-3.1-pro-preview",
      "provider_tag": "google-ai-studio/flex",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 8,
      "strict_completion_rate": 0.6666666666666666,
      "strict_conservative_cost_exposure_usd": 0.9922683333333333,
      "strict_epicure_calls": 10,
      "strict_epicure_successful_calls": 9,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.39062208887279953,
      "strict_wilson_upper_95": 0.8618799089087867,
      "total_complete_pairs": 14,
      "total_conservative_cost_exposure_usd": 1.2971183333333334,
      "total_epicure_calls": 16,
      "total_epicure_successful_calls": 15,
      "total_scheduled_pairs": 20
    },
    {
      "canonical_model_slug": "google/gemini-3.6-flash-20260721",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": 4.1666666666666625,
      "display_name": "Gemini 3.6 Flash",
      "high_resource_complete_pairs": 7,
      "high_resource_completion_rate": 0.875,
      "high_resource_conservative_cost_exposure_usd": 0.149434,
      "high_resource_epicure_calls": 8,
      "high_resource_epicure_successful_calls": 7,
      "high_resource_scheduled_pairs": 8,
      "high_resource_wilson_lower_95": 0.5291118177871464,
      "high_resource_wilson_upper_95": 0.9775825085499433,
      "manifest_ordinal": 6,
      "minimum_eight_complete_pairs": true,
      "model_id": "google/gemini-3.6-flash",
      "provider_tag": "google-ai-studio/flex",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 10,
      "strict_completion_rate": 0.8333333333333334,
      "strict_conservative_cost_exposure_usd": 0.204509,
      "strict_epicure_calls": 12,
      "strict_epicure_successful_calls": 11,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.5519691377470266,
      "strict_wilson_upper_95": 0.9530348578161463,
      "total_complete_pairs": 17,
      "total_conservative_cost_exposure_usd": 0.353943,
      "total_epicure_calls": 20,
      "total_epicure_successful_calls": 18,
      "total_scheduled_pairs": 20
    },
    {
      "canonical_model_slug": "x-ai/grok-4.5-20260708",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": 66.66666666666666,
      "display_name": "Grok 4.5",
      "high_resource_complete_pairs": 8,
      "high_resource_completion_rate": 0.6666666666666666,
      "high_resource_conservative_cost_exposure_usd": 0.512772,
      "high_resource_epicure_calls": 22,
      "high_resource_epicure_successful_calls": 20,
      "high_resource_scheduled_pairs": 12,
      "high_resource_wilson_lower_95": 0.39062208887279953,
      "high_resource_wilson_upper_95": 0.8618799089087867,
      "manifest_ordinal": 7,
      "minimum_eight_complete_pairs": true,
      "model_id": "x-ai/grok-4.5",
      "provider_tag": "xai/zdr",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 0,
      "strict_completion_rate": 0.0,
      "strict_conservative_cost_exposure_usd": 0.365594,
      "strict_epicure_calls": 0,
      "strict_epicure_successful_calls": 0,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 2.7755575615628914e-17,
      "strict_wilson_upper_95": 0.24249400665524085,
      "total_complete_pairs": 8,
      "total_conservative_cost_exposure_usd": 0.878366,
      "total_epicure_calls": 22,
      "total_epicure_successful_calls": 20,
      "total_scheduled_pairs": 24
    },
    {
      "canonical_model_slug": "k3",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": 12.5,
      "display_name": "Kimi K3",
      "high_resource_complete_pairs": 7,
      "high_resource_completion_rate": 0.875,
      "high_resource_conservative_cost_exposure_usd": 13.789952,
      "high_resource_epicure_calls": 13,
      "high_resource_epicure_successful_calls": 10,
      "high_resource_scheduled_pairs": 8,
      "high_resource_wilson_lower_95": 0.5291118177871464,
      "high_resource_wilson_upper_95": 0.9775825085499433,
      "manifest_ordinal": 8,
      "minimum_eight_complete_pairs": true,
      "model_id": "moonshotai/kimi-k3",
      "provider_tag": "kimi-code-direct",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 9,
      "strict_completion_rate": 0.75,
      "strict_conservative_cost_exposure_usd": 11.661807,
      "strict_epicure_calls": 21,
      "strict_epicure_successful_calls": 17,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.46769466506643426,
      "strict_wilson_upper_95": 0.9110583316059453,
      "total_complete_pairs": 16,
      "total_conservative_cost_exposure_usd": 25.451759,
      "total_epicure_calls": 34,
      "total_epicure_successful_calls": 27,
      "total_scheduled_pairs": 20
    },
    {
      "canonical_model_slug": "z-ai/glm-5.2-20260616",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": 20.833333333333332,
      "display_name": "GLM 5.2",
      "high_resource_complete_pairs": 5,
      "high_resource_completion_rate": 0.625,
      "high_resource_conservative_cost_exposure_usd": 0.172205,
      "high_resource_epicure_calls": 19,
      "high_resource_epicure_successful_calls": 14,
      "high_resource_scheduled_pairs": 8,
      "high_resource_wilson_lower_95": 0.30574239460262737,
      "high_resource_wilson_upper_95": 0.8631557141764026,
      "manifest_ordinal": 9,
      "minimum_eight_complete_pairs": true,
      "model_id": "z-ai/glm-5.2",
      "provider_tag": "deepinfra/fp4",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 5,
      "strict_completion_rate": 0.4166666666666667,
      "strict_conservative_cost_exposure_usd": 0.194648,
      "strict_epicure_calls": 16,
      "strict_epicure_successful_calls": 14,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.19326031365875665,
      "strict_wilson_upper_95": 0.6804886874504503,
      "total_complete_pairs": 10,
      "total_conservative_cost_exposure_usd": 0.366853,
      "total_epicure_calls": 35,
      "total_epicure_successful_calls": 28,
      "total_scheduled_pairs": 20
    },
    {
      "canonical_model_slug": "deepseek/deepseek-v4-pro-20260423",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": 20.833333333333332,
      "display_name": "DeepSeek V4 Pro",
      "high_resource_complete_pairs": 5,
      "high_resource_completion_rate": 0.625,
      "high_resource_conservative_cost_exposure_usd": 0.9554422,
      "high_resource_epicure_calls": 19,
      "high_resource_epicure_successful_calls": 11,
      "high_resource_scheduled_pairs": 8,
      "high_resource_wilson_lower_95": 0.30574239460262737,
      "high_resource_wilson_upper_95": 0.8631557141764026,
      "manifest_ordinal": 10,
      "minimum_eight_complete_pairs": true,
      "model_id": "deepseek/deepseek-v4-pro",
      "provider_tag": "cloudflare",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 5,
      "strict_completion_rate": 0.4166666666666667,
      "strict_conservative_cost_exposure_usd": 0.77478866,
      "strict_epicure_calls": 17,
      "strict_epicure_successful_calls": 13,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.19326031365875665,
      "strict_wilson_upper_95": 0.6804886874504503,
      "total_complete_pairs": 10,
      "total_conservative_cost_exposure_usd": 1.73023086,
      "total_epicure_calls": 36,
      "total_epicure_successful_calls": 24,
      "total_scheduled_pairs": 20
    },
    {
      "canonical_model_slug": "deepseek/deepseek-v4-flash-20260731",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": -4.1666666666666625,
      "display_name": "DeepSeek V4 Flash",
      "high_resource_complete_pairs": 7,
      "high_resource_completion_rate": 0.875,
      "high_resource_conservative_cost_exposure_usd": 0.019376,
      "high_resource_epicure_calls": 23,
      "high_resource_epicure_successful_calls": 20,
      "high_resource_scheduled_pairs": 8,
      "high_resource_wilson_lower_95": 0.5291118177871464,
      "high_resource_wilson_upper_95": 0.9775825085499433,
      "manifest_ordinal": 11,
      "minimum_eight_complete_pairs": true,
      "model_id": "deepseek/deepseek-v4-flash-0731",
      "provider_tag": "deepinfra/fp4",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 11,
      "strict_completion_rate": 0.9166666666666666,
      "strict_conservative_cost_exposure_usd": 0.028436,
      "strict_epicure_calls": 28,
      "strict_epicure_successful_calls": 22,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.646120088858883,
      "strict_wilson_upper_95": 0.9851349055950829,
      "total_complete_pairs": 18,
      "total_conservative_cost_exposure_usd": 0.047812,
      "total_epicure_calls": 51,
      "total_epicure_successful_calls": 42,
      "total_scheduled_pairs": 20
    },
    {
      "canonical_model_slug": "minimax/minimax-m3-20260531",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": 41.66666666666667,
      "display_name": "MiniMax M3",
      "high_resource_complete_pairs": 7,
      "high_resource_completion_rate": 0.5833333333333334,
      "high_resource_conservative_cost_exposure_usd": 0.156687,
      "high_resource_epicure_calls": 27,
      "high_resource_epicure_successful_calls": 10,
      "high_resource_scheduled_pairs": 12,
      "high_resource_wilson_lower_95": 0.31951131254954973,
      "high_resource_wilson_upper_95": 0.8067396863412435,
      "manifest_ordinal": 12,
      "minimum_eight_complete_pairs": true,
      "model_id": "minimax/minimax-m3",
      "provider_tag": "morph",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 2,
      "strict_completion_rate": 0.16666666666666666,
      "strict_conservative_cost_exposure_usd": 0.12928,
      "strict_epicure_calls": 9,
      "strict_epicure_successful_calls": 4,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.04696514218385378,
      "strict_wilson_upper_95": 0.4480308622529734,
      "total_complete_pairs": 9,
      "total_conservative_cost_exposure_usd": 0.28596699999999997,
      "total_epicure_calls": 36,
      "total_epicure_successful_calls": 14,
      "total_scheduled_pairs": 24
    },
    {
      "canonical_model_slug": "nvidia/nemotron-3-ultra-550b-a55b-20260604",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": 8.333333333333337,
      "display_name": "Nemotron 3 Ultra",
      "high_resource_complete_pairs": 6,
      "high_resource_completion_rate": 0.75,
      "high_resource_conservative_cost_exposure_usd": 0.236177,
      "high_resource_epicure_calls": 15,
      "high_resource_epicure_successful_calls": 11,
      "high_resource_scheduled_pairs": 8,
      "high_resource_wilson_lower_95": 0.40927543031016883,
      "high_resource_wilson_upper_95": 0.9285207872478909,
      "manifest_ordinal": 13,
      "minimum_eight_complete_pairs": true,
      "model_id": "nvidia/nemotron-3-ultra-550b-a55b",
      "provider_tag": "together",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 8,
      "strict_completion_rate": 0.6666666666666666,
      "strict_conservative_cost_exposure_usd": 0.674427,
      "strict_epicure_calls": 11,
      "strict_epicure_successful_calls": 10,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.39062208887279953,
      "strict_wilson_upper_95": 0.8618799089087867,
      "total_complete_pairs": 14,
      "total_conservative_cost_exposure_usd": 0.910604,
      "total_epicure_calls": 26,
      "total_epicure_successful_calls": 21,
      "total_scheduled_pairs": 20
    },
    {
      "canonical_model_slug": "mistralai/mistral-medium-3.5-20260430",
      "cost_display_status": "known_usd_exposure",
      "descriptive_difference_percentage_points": -8.333333333333332,
      "display_name": "Mistral Medium 3.5",
      "high_resource_complete_pairs": 4,
      "high_resource_completion_rate": 0.25,
      "high_resource_conservative_cost_exposure_usd": 1.756442,
      "high_resource_epicure_calls": 5,
      "high_resource_epicure_successful_calls": 5,
      "high_resource_scheduled_pairs": 16,
      "high_resource_wilson_lower_95": 0.10182067491213048,
      "high_resource_wilson_upper_95": 0.49498316535508774,
      "manifest_ordinal": 14,
      "minimum_eight_complete_pairs": true,
      "model_id": "mistralai/mistral-medium-3-5",
      "provider_tag": "mistral",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": 4,
      "strict_completion_rate": 0.3333333333333333,
      "strict_conservative_cost_exposure_usd": 1.063717,
      "strict_epicure_calls": 4,
      "strict_epicure_successful_calls": 4,
      "strict_observed": true,
      "strict_scheduled_pairs": 12,
      "strict_wilson_lower_95": 0.13812009109121315,
      "strict_wilson_upper_95": 0.6093779111272004,
      "total_complete_pairs": 8,
      "total_conservative_cost_exposure_usd": 2.8201590000000003,
      "total_epicure_calls": 9,
      "total_epicure_successful_calls": 9,
      "total_scheduled_pairs": 28
    },
    {
      "canonical_model_slug": "command-a-plus-05-2026",
      "cost_display_status": "provider_charge_unavailable",
      "descriptive_difference_percentage_points": null,
      "display_name": "Command A Plus",
      "high_resource_complete_pairs": 12,
      "high_resource_completion_rate": 1.0,
      "high_resource_conservative_cost_exposure_usd": null,
      "high_resource_epicure_calls": 15,
      "high_resource_epicure_successful_calls": 12,
      "high_resource_scheduled_pairs": 12,
      "high_resource_wilson_lower_95": 0.7575059933447592,
      "high_resource_wilson_upper_95": 1.0,
      "manifest_ordinal": 15,
      "minimum_eight_complete_pairs": true,
      "model_id": "cohere/command-a-plus-05-2026",
      "provider_tag": "cohere-direct",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": null,
      "strict_completion_rate": null,
      "strict_conservative_cost_exposure_usd": null,
      "strict_epicure_calls": 0,
      "strict_epicure_successful_calls": 0,
      "strict_observed": false,
      "strict_scheduled_pairs": null,
      "strict_wilson_lower_95": null,
      "strict_wilson_upper_95": null,
      "total_complete_pairs": 12,
      "total_conservative_cost_exposure_usd": null,
      "total_epicure_calls": 15,
      "total_epicure_successful_calls": 12,
      "total_scheduled_pairs": 12
    },
    {
      "canonical_model_slug": "command-a-reasoning-08-2025",
      "cost_display_status": "provider_charge_unavailable",
      "descriptive_difference_percentage_points": null,
      "display_name": "Command A Reasoning",
      "high_resource_complete_pairs": 8,
      "high_resource_completion_rate": 0.6666666666666666,
      "high_resource_conservative_cost_exposure_usd": null,
      "high_resource_epicure_calls": 19,
      "high_resource_epicure_successful_calls": 18,
      "high_resource_scheduled_pairs": 12,
      "high_resource_wilson_lower_95": 0.39062208887279953,
      "high_resource_wilson_upper_95": 0.8618799089087867,
      "manifest_ordinal": 16,
      "minimum_eight_complete_pairs": true,
      "model_id": "cohere/command-a-reasoning-08-2025",
      "provider_tag": "cohere-direct",
      "quality_judgments": 0,
      "rank_eligible": false,
      "strict_complete_pairs": null,
      "strict_completion_rate": null,
      "strict_conservative_cost_exposure_usd": null,
      "strict_epicure_calls": 0,
      "strict_epicure_successful_calls": 0,
      "strict_observed": false,
      "strict_scheduled_pairs": null,
      "strict_wilson_lower_95": null,
      "strict_wilson_upper_95": null,
      "total_complete_pairs": 8,
      "total_conservative_cost_exposure_usd": null,
      "total_epicure_calls": 19,
      "total_epicure_successful_calls": 18,
      "total_scheduled_pairs": 12
    }
  ],
  "official": false,
  "quality_judgments": 0,
  "quality_ranking": false,
  "rank_eligible": false,
  "schema_version": "flavourbench-frontier-protocol-sensitivity-v1",
  "status": "verified_real_protocol_sensitivity",
  "strict": {
    "complete_pairs": 101,
    "completed_response_arms": 262,
    "distinct_tasks": 12,
    "epicure_calls": 215,
    "epicure_successful_calls": 173,
    "execution_policy_sha256": "5c02a464cf7c6632ea35f88ca9bbf10a527a976d2ceb3a0bd1d62a40cbe1e6c4",
    "known_conservative_exposure_subtotal_usd": 26.545659993333334,
    "provider_charge_complete": true,
    "provider_generation_ids": 964,
    "scheduled_pairs": 168,
    "source_artifact_sha256": "52c1f677235f30aae4d45f99f8b29cbf0d557928ecf311b2a07d89b63a7e8be0",
    "task_set_sha256": "a012a9c968cc697b644298ef039789d73196e48e4acfddca0156126b15b47e01",
    "unpriced_model_ids": []
  },
  "synthetic_tasks": 0,
  "task_sets_disjoint": true
}
