{"schema":"hive-bench-results-site","schema_version":14,"corpus_version":"v2-ce + v3-mixed-workflows-followup-20260713 + v3-production-review-panels-20260723 + v3-deepseek-v4-0813-20260813-r2 + v3-pi-ox-alpha-high-20260825-r3 + v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed","generated_at":"2026-08-27T22:21:35Z","status":"preliminary_complete","coverage":{"cells":90,"expected_cells":90,"tasks":6,"candidates":15},"efficiency_accounting":{"source":"campaign-serialized generation telemetry","parser":"HiveBench::TokenReport","pricing_version":"2026-06-usual + 2026-08-12-openrouter","priced_cells":52,"total_cells":90,"judge_usage_included":false,"excluded_cumulative_event_types":["result","system"],"pi_usage_event":"assistant message_end","codex_cached_input_split":true,"token_total_buckets":["input","output","cache_read","cache_write"],"followup_priced_cells":5,"followup_timed_cells":26,"known_partial_cost_cells":24,"deepseek_priced_cells":11,"deepseek_timed_cells":11,"ox_alpha_priced_cells":6,"ox_alpha_token_cells":12,"ox_alpha_timed_cells":9},"discussion_adjusted":{"status":"diagnostic","source_campaign_ids":["v2-ce","v3-mixed-workflows-three-seed-20260713","v3-production-review-panels-three-seed-20260723","v3-deepseek-v4-0813-20260813-r2","v3-pi-ox-alpha-high-20260825-r3","v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed"],"description":"A separate one-shot diagnostic covering all six active campaigns. The v2-ce campaign reused its exact published independent verdicts and rationales recovered from exact local provider sessions for round one; the five active three-sample campaigns freshly re-graded round one. In every active campaign, each judge then saw the other judge's verdict anonymously, argued the strongest evidence-based case that its own view was wrong, and held or revised. Discussion-final scores do not replace the independent leaderboard scores because shared verdicts can introduce anchoring or convergence pressure.","coverage":{"cells":90,"expected_cells":90,"judge_decisions":180,"expected_judge_decisions":180,"fully_adjusted_cells":90,"missing":[]},"summary":{"cells":90,"mean_revision_by_judge":{"fable-5":-0.829,"gpt-5.6-sol":-0.084},"mean_abs_revision_by_judge":{"fable-5":0.842,"gpt-5.6-sol":0.3},"mean_spread_before":2.082,"mean_spread_after":1.251},"candidates":[{"id":"all-codex-5.6-sol-xhigh","fable":{"mean":6.25,"sample":6,"total":6},"sol":{"mean":5.867,"sample":6,"total":6},"combined":{"mean":6.058,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":7.0,"final":7.0,"delta":0.0,"revised":false},"sol":{"initial":7.0,"final":7.0,"delta":0.0,"revised":false}},"web-install":{"fable":{"initial":7.0,"final":6.5,"delta":-0.5,"revised":true},"sol":{"initial":5.5,"final":5.5,"delta":0.0,"revised":false}},"install":{"fable":{"initial":5.5,"final":4.0,"delta":-1.5,"revised":true},"sol":{"initial":3.5,"final":3.5,"delta":0.0,"revised":false}},"fix-tmux":{"fable":{"initial":9.0,"final":8.5,"delta":-0.5,"revised":true},"sol":{"initial":8.5,"final":8.5,"delta":0.0,"revised":false}},"fix-review":{"fable":{"initial":6.0,"final":6.0,"delta":0.0,"revised":false},"sol":{"initial":7.1,"final":6.2,"delta":-0.9,"revised":true}},"daemon":{"fable":{"initial":7.0,"final":5.5,"delta":-1.5,"revised":true},"sol":{"initial":5.5,"final":4.5,"delta":-1.0,"revised":true}}}},{"id":"opus-plan->codex-exec-xhigh","fable":{"mean":5.583,"sample":6,"total":6},"sol":{"mean":4.95,"sample":6,"total":6},"combined":{"mean":5.267,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":6.0,"final":5.5,"delta":-0.5,"revised":true},"sol":{"initial":5.0,"final":5.0,"delta":0.0,"revised":false}},"web-install":{"fable":{"initial":5.0,"final":4.0,"delta":-1.0,"revised":true},"sol":{"initial":3.0,"final":4.0,"delta":1.0,"revised":true}},"install":{"fable":{"initial":6.5,"final":6.0,"delta":-0.5,"revised":true},"sol":{"initial":4.5,"final":4.5,"delta":0.0,"revised":false}},"fix-tmux":{"fable":{"initial":9.0,"final":9.0,"delta":0.0,"revised":false},"sol":{"initial":8.7,"final":8.7,"delta":0.0,"revised":false}},"fix-review":{"fable":{"initial":4.0,"final":3.5,"delta":-0.5,"revised":true},"sol":{"initial":3.0,"final":3.0,"delta":0.0,"revised":false}},"daemon":{"fable":{"initial":6.5,"final":5.5,"delta":-1.0,"revised":true},"sol":{"initial":4.5,"final":4.5,"delta":0.0,"revised":false}}}},{"id":"all-opus-4.8","fable":{"mean":5.5,"sample":6,"total":6},"sol":{"mean":4.167,"sample":6,"total":6},"combined":{"mean":4.833,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":6.0,"final":5.5,"delta":-0.5,"revised":true},"sol":{"initial":5.0,"final":5.0,"delta":0.0,"revised":false}},"web-install":{"fable":{"initial":4.5,"final":4.0,"delta":-0.5,"revised":true},"sol":{"initial":2.5,"final":2.5,"delta":0.0,"revised":false}},"install":{"fable":{"initial":6.0,"final":6.0,"delta":0.0,"revised":false},"sol":{"initial":2.5,"final":2.5,"delta":0.0,"revised":false}},"fix-tmux":{"fable":{"initial":8.0,"final":7.5,"delta":-0.5,"revised":true},"sol":{"initial":7.5,"final":7.5,"delta":0.0,"revised":false}},"fix-review":{"fable":{"initial":5.0,"final":4.0,"delta":-1.0,"revised":true},"sol":{"initial":3.0,"final":2.5,"delta":-0.5,"revised":true}},"daemon":{"fable":{"initial":7.0,"final":6.0,"delta":-1.0,"revised":true},"sol":{"initial":5.0,"final":5.0,"delta":0.0,"revised":false}}}},{"id":"all-grok-4.5","fable":{"mean":4.883,"sample":6,"total":6},"sol":{"mean":3.883,"sample":6,"total":6},"combined":{"mean":4.383,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":5.5,"final":5.0,"delta":-0.5,"revised":true},"sol":{"initial":5.1,"final":4.8,"delta":-0.3,"revised":true}},"web-install":{"fable":{"initial":6.0,"final":5.0,"delta":-1.0,"revised":true},"sol":{"initial":4.0,"final":3.5,"delta":-0.5,"revised":true}},"install":{"fable":{"initial":6.5,"final":3.0,"delta":-3.5,"revised":true},"sol":{"initial":2.0,"final":2.0,"delta":0.0,"revised":false}},"fix-tmux":{"fable":{"initial":7.5,"final":6.8,"delta":-0.7,"revised":true},"sol":{"initial":6.2,"final":5.5,"delta":-0.7,"revised":true}},"fix-review":{"fable":{"initial":5.0,"final":3.5,"delta":-1.5,"revised":true},"sol":{"initial":3.0,"final":3.0,"delta":0.0,"revised":false}},"daemon":{"fable":{"initial":7.5,"final":6.0,"delta":-1.5,"revised":true},"sol":{"initial":5.0,"final":4.5,"delta":-0.5,"revised":true}}}},{"id":"all-codex-xhigh","fable":{"mean":4.0,"sample":6,"total":6},"sol":{"mean":3.45,"sample":6,"total":6},"combined":{"mean":3.725,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":5.5,"final":5.0,"delta":-0.5,"revised":true},"sol":{"initial":4.5,"final":4.5,"delta":0.0,"revised":false}},"web-install":{"fable":{"initial":3.0,"final":3.0,"delta":0.0,"revised":false},"sol":{"initial":3.5,"final":2.5,"delta":-1.0,"revised":true}},"install":{"fable":{"initial":5.0,"final":3.5,"delta":-1.5,"revised":true},"sol":{"initial":2.5,"final":2.5,"delta":0.0,"revised":false}},"fix-tmux":{"fable":{"initial":5.0,"final":5.0,"delta":0.0,"revised":false},"sol":{"initial":6.5,"final":5.5,"delta":-1.0,"revised":true}},"fix-review":{"fable":{"initial":3.5,"final":3.0,"delta":-0.5,"revised":true},"sol":{"initial":2.0,"final":1.5,"delta":-0.5,"revised":true}},"daemon":{"fable":{"initial":6.0,"final":4.5,"delta":-1.5,"revised":true},"sol":{"initial":4.2,"final":4.2,"delta":0.0,"revised":false}}}},{"id":"all-glm-5.2","fable":{"mean":4.25,"sample":6,"total":6},"sol":{"mean":3.383,"sample":6,"total":6},"combined":{"mean":3.817,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":6.5,"final":6.0,"delta":-0.5,"revised":true},"sol":{"initial":5.5,"final":5.0,"delta":-0.5,"revised":true}},"web-install":{"fable":{"initial":3.0,"final":3.0,"delta":0.0,"revised":false},"sol":{"initial":3.0,"final":3.0,"delta":0.0,"revised":false}},"install":{"fable":{"initial":6.0,"final":4.0,"delta":-2.0,"revised":true},"sol":{"initial":2.0,"final":2.5,"delta":0.5,"revised":true}},"fix-tmux":{"fable":{"initial":7.0,"final":7.5,"delta":0.5,"revised":true},"sol":{"initial":7.2,"final":6.8,"delta":-0.4,"revised":true}},"fix-review":{"fable":{"initial":1.5,"final":1.0,"delta":-0.5,"revised":true},"sol":{"initial":1.0,"final":0.5,"delta":-0.5,"revised":true}},"daemon":{"fable":{"initial":5.0,"final":4.0,"delta":-1.0,"revised":true},"sol":{"initial":3.0,"final":2.5,"delta":-0.5,"revised":true}}}},{"id":"fable-plan->grok-exec-sol-review","fable":{"mean":6.583,"sample":6,"total":6},"sol":{"mean":5.517,"sample":6,"total":6},"combined":{"mean":6.05,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":8.5,"final":8,"delta":-0.5,"revised":true},"sol":{"initial":8,"final":8,"delta":0,"revised":false}},"web-install":{"fable":{"initial":5,"final":4.5,"delta":-0.5,"revised":true},"sol":{"initial":4,"final":3.5,"delta":-0.5,"revised":true}},"install":{"fable":{"initial":7,"final":4,"delta":-3,"revised":true},"sol":{"initial":3,"final":2.5,"delta":-0.5,"revised":true}},"fix-tmux":{"fable":{"initial":9,"final":9,"delta":0,"revised":false},"sol":{"initial":7.5,"final":7.5,"delta":0,"revised":false}},"fix-review":{"fable":{"initial":8,"final":7.5,"delta":-0.5,"revised":true},"sol":{"initial":6.5,"final":6,"delta":-0.5,"revised":true}},"daemon":{"fable":{"initial":7,"final":6.5,"delta":-0.5,"revised":true},"sol":{"initial":5.8,"final":5.6,"delta":-0.2,"revised":true}}}},{"id":"sol-plan->grok-exec-sol-review","fable":{"mean":7.383,"sample":6,"total":6},"sol":{"mean":6.433,"sample":6,"total":6},"combined":{"mean":6.908,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":9,"final":8.8,"delta":-0.2,"revised":true},"sol":{"initial":8.2,"final":8.5,"delta":0.3,"revised":true}},"web-install":{"fable":{"initial":7.5,"final":7,"delta":-0.5,"revised":true},"sol":{"initial":5.8,"final":5.8,"delta":0,"revised":false}},"install":{"fable":{"initial":8,"final":6.5,"delta":-1.5,"revised":true},"sol":{"initial":4.8,"final":5.3,"delta":0.5,"revised":true}},"fix-tmux":{"fable":{"initial":9,"final":8.5,"delta":-0.5,"revised":true},"sol":{"initial":8,"final":8.5,"delta":0.5,"revised":true}},"fix-review":{"fable":{"initial":7.5,"final":5.5,"delta":-2,"revised":true},"sol":{"initial":4.5,"final":4,"delta":-0.5,"revised":true}},"daemon":{"fable":{"initial":8.5,"final":8,"delta":-0.5,"revised":true},"sol":{"initial":6.2,"final":6.5,"delta":0.3,"revised":true}}}},{"id":"sol-plan->terra-exec-sol-review","fable":{"mean":7.45,"sample":6,"total":6},"sol":{"mean":5.483,"sample":6,"total":6},"combined":{"mean":6.467,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":9.5,"final":9.2,"delta":-0.3,"revised":true},"sol":{"initial":8.6,"final":8.9,"delta":0.3,"revised":true}},"web-install":{"fable":{"initial":5,"final":4.5,"delta":-0.5,"revised":true},"sol":{"initial":4,"final":3.5,"delta":-0.5,"revised":true}},"install":{"fable":{"initial":7.5,"final":6,"delta":-1.5,"revised":true},"sol":{"initial":4.2,"final":4.2,"delta":0,"revised":false}},"fix-tmux":{"fable":{"initial":9,"final":9,"delta":0,"revised":false},"sol":{"initial":6,"final":6,"delta":0,"revised":false}},"fix-review":{"fable":{"initial":9,"final":7.5,"delta":-1.5,"revised":true},"sol":{"initial":7.1,"final":6.8,"delta":-0.3,"revised":true}},"daemon":{"fable":{"initial":8.5,"final":8.5,"delta":0,"revised":false},"sol":{"initial":3.5,"final":3.5,"delta":0,"revised":false}}}},{"id":"sol-plan->terra-exec-grok-review","fable":{"mean":6.517,"sample":6,"total":6},"sol":{"mean":4.917,"sample":6,"total":6},"combined":{"mean":5.717,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":8.5,"final":8.6,"delta":0.1,"revised":true},"sol":{"initial":8.7,"final":8.5,"delta":-0.2,"revised":true}},"web-install":{"fable":{"initial":6.5,"final":5.0,"delta":-1.5,"revised":true},"sol":{"initial":3.0,"final":3.0,"delta":0.0,"revised":false}},"install":{"fable":{"initial":7.0,"final":6.0,"delta":-1.0,"revised":true},"sol":{"initial":3.0,"final":2.5,"delta":-0.5,"revised":true}},"fix-tmux":{"fable":{"initial":8.5,"final":7.5,"delta":-1.0,"revised":true},"sol":{"initial":7.0,"final":7.0,"delta":0.0,"revised":false}},"fix-review":{"fable":{"initial":8.5,"final":7.5,"delta":-1.0,"revised":true},"sol":{"initial":5.5,"final":5.0,"delta":-0.5,"revised":true}},"daemon":{"fable":{"initial":6.5,"final":4.5,"delta":-2.0,"revised":true},"sol":{"initial":4.0,"final":3.5,"delta":-0.5,"revised":true}}}},{"id":"sol-plan@xhigh->exec@high+grok-review","fable":{"mean":7.917,"sample":6,"total":6},"sol":{"mean":7.133,"sample":6,"total":6},"combined":{"mean":7.525,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":9.0,"final":8.5,"delta":-0.5,"revised":true},"sol":{"initial":8.5,"final":8.5,"delta":0.0,"revised":false}},"web-install":{"fable":{"initial":7.5,"final":7.0,"delta":-0.5,"revised":true},"sol":{"initial":6.5,"final":6.5,"delta":0.0,"revised":false}},"install":{"fable":{"initial":8.0,"final":6.0,"delta":-2.0,"revised":true},"sol":{"initial":4.2,"final":4.0,"delta":-0.2,"revised":true}},"fix-tmux":{"fable":{"initial":9.0,"final":9.0,"delta":0.0,"revised":false},"sol":{"initial":9.2,"final":9.2,"delta":0.0,"revised":false}},"fix-review":{"fable":{"initial":9.0,"final":8.5,"delta":-0.5,"revised":true},"sol":{"initial":6.8,"final":7.1,"delta":0.3,"revised":true}},"daemon":{"fable":{"initial":9.0,"final":8.5,"delta":-0.5,"revised":true},"sol":{"initial":6.5,"final":7.5,"delta":1.0,"revised":true}}}},{"id":"all-deepseek-v4-pro-0813@xhigh","fable":{"mean":6.833,"sample":6,"total":6},"sol":{"mean":4.95,"sample":6,"total":6},"combined":{"mean":5.892,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":9.0,"final":9.0,"delta":0.0,"revised":false},"sol":{"initial":9.2,"final":9.2,"delta":0.0,"revised":false}},"web-install":{"fable":{"initial":7.5,"final":7.5,"delta":0.0,"revised":false},"sol":{"initial":3.0,"final":3.5,"delta":0.5,"revised":true}},"install":{"fable":{"initial":7.0,"final":6.0,"delta":-1.0,"revised":true},"sol":{"initial":3.0,"final":3.0,"delta":0.0,"revised":false}},"fix-tmux":{"fable":{"initial":5.0,"final":5.0,"delta":0.0,"revised":false},"sol":{"initial":6.0,"final":5.0,"delta":-1.0,"revised":true}},"fix-review":{"fable":{"initial":8.0,"final":6.0,"delta":-2.0,"revised":true},"sol":{"initial":4.2,"final":4.0,"delta":-0.2,"revised":true}},"daemon":{"fable":{"initial":8.0,"final":7.5,"delta":-0.5,"revised":true},"sol":{"initial":5.0,"final":5.0,"delta":0.0,"revised":false}}}},{"id":"deepseek-pro@xhigh->flash@xhigh-exec+pro-review","fable":{"mean":6.75,"sample":6,"total":6},"sol":{"mean":4.817,"sample":6,"total":6},"combined":{"mean":5.783,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":9.0,"final":9.0,"delta":0.0,"revised":false},"sol":{"initial":9.5,"final":9.4,"delta":-0.1,"revised":true}},"web-install":{"fable":{"initial":7.5,"final":7.0,"delta":-0.5,"revised":true},"sol":{"initial":3.0,"final":3.0,"delta":0.0,"revised":false}},"install":{"fable":{"initial":7.0,"final":5.0,"delta":-2.0,"revised":true},"sol":{"initial":2.0,"final":2.0,"delta":0.0,"revised":false}},"fix-tmux":{"fable":{"initial":8.5,"final":8.0,"delta":-0.5,"revised":true},"sol":{"initial":6.5,"final":6.5,"delta":0.0,"revised":false}},"fix-review":{"fable":{"initial":5.0,"final":4.5,"delta":-0.5,"revised":true},"sol":{"initial":2.5,"final":2.5,"delta":0.0,"revised":false}},"daemon":{"fable":{"initial":8.0,"final":7.0,"delta":-1.0,"revised":true},"sol":{"initial":5.5,"final":5.5,"delta":0.0,"revised":false}}}},{"id":"all-ox-alpha@high","fable":{"mean":6.333,"sample":6,"total":6},"sol":{"mean":5.05,"sample":6,"total":6},"combined":{"mean":5.692,"sample":6,"total":6},"cells":{"install":{"fable":{"initial":7.0,"final":4.0,"delta":-3.0,"revised":true},"sol":{"initial":3.1,"final":2.8,"delta":-0.3,"revised":true}},"fix-tmux":{"fable":{"initial":8.0,"final":7.5,"delta":-0.5,"revised":true},"sol":{"initial":7.0,"final":6.5,"delta":-0.5,"revised":true}},"daemon":{"fable":{"initial":7.0,"final":6.5,"delta":-0.5,"revised":true},"sol":{"initial":5.0,"final":5.0,"delta":0.0,"revised":false}},"add-i-key":{"fable":{"initial":8.5,"final":8.0,"delta":-0.5,"revised":true},"sol":{"initial":8.0,"final":7.5,"delta":-0.5,"revised":true}},"web-install":{"fable":{"initial":8.0,"final":7.5,"delta":-0.5,"revised":true},"sol":{"initial":2.0,"final":6.5,"delta":4.5,"revised":true}},"fix-review":{"fable":{"initial":7.0,"final":4.5,"delta":-2.5,"revised":true},"sol":{"initial":2.0,"final":2.0,"delta":0.0,"revised":false}}}},{"id":"all-ox-alpha@max","fable":{"mean":6.5,"sample":6,"total":6},"sol":{"mean":4.467,"sample":6,"total":6},"combined":{"mean":5.483,"sample":6,"total":6},"cells":{"add-i-key":{"fable":{"initial":8.5,"final":8.0,"delta":-0.5,"revised":true},"sol":{"initial":7.5,"final":7.5,"delta":0.0,"revised":false}},"web-install":{"fable":{"initial":7.5,"final":5.0,"delta":-2.5,"revised":true},"sol":{"initial":2.5,"final":2.5,"delta":0.0,"revised":false}},"install":{"fable":{"initial":6.0,"final":4.0,"delta":-2.0,"revised":true},"sol":{"initial":2.5,"final":2.5,"delta":0.0,"revised":false}},"fix-tmux":{"fable":{"initial":9.0,"final":9.0,"delta":0.0,"revised":false},"sol":{"initial":8.8,"final":8.8,"delta":0.0,"revised":false}},"fix-review":{"fable":{"initial":7.0,"final":5.0,"delta":-2.0,"revised":true},"sol":{"initial":3.0,"final":2.5,"delta":-0.5,"revised":true}},"daemon":{"fable":{"initial":8.5,"final":8.0,"delta":-0.5,"revised":true},"sol":{"initial":3.5,"final":3.0,"delta":-0.5,"revised":true}}}}]},"primary_judges":[{"id":"gpt-5.6-sol","label":"GPT-5.6 Sol","reasoning_effort":"varies by campaign","samples_per_cell":null,"rows":[{"id":"all-codex-5.6-sol-xhigh","label":"GPT-5.6 Sol xhigh","mean":6.183,"scored":6,"total":6,"same_family":true,"campaign_id":"v2-ce","reasoning_effort":"xhigh","samples_per_cell":1},{"id":"opus-plan->codex-exec-xhigh","label":"Opus plan → Codex 5.5 xhigh","mean":4.783,"scored":6,"total":6,"same_family":true,"campaign_id":"v2-ce","reasoning_effort":"xhigh","samples_per_cell":1},{"id":"all-opus-4.8","label":"Opus 4.8","mean":4.25,"scored":6,"total":6,"same_family":false,"campaign_id":"v2-ce","reasoning_effort":"xhigh","samples_per_cell":1},{"id":"all-grok-4.5","label":"Grok 4.5 xhigh","mean":4.217,"scored":6,"total":6,"same_family":false,"campaign_id":"v2-ce","reasoning_effort":"xhigh","samples_per_cell":1},{"id":"all-codex-xhigh","label":"Codex 5.5 xhigh","mean":3.867,"scored":6,"total":6,"same_family":true,"campaign_id":"v2-ce","reasoning_effort":"xhigh","samples_per_cell":1},{"id":"all-glm-5.2","label":"GLM 5.2","mean":3.617,"scored":6,"total":6,"same_family":false,"campaign_id":"v2-ce","reasoning_effort":"xhigh","samples_per_cell":1},{"id":"fable-plan->grok-exec-sol-review","label":"Fable plan → Grok execute → Sol review","mean":4.667,"scored":6,"total":6,"same_family":true,"campaign_id":"v3-mixed-workflows-three-seed-20260713","reasoning_effort":"ultra","samples_per_cell":3},{"id":"sol-plan->grok-exec-sol-review","label":"Sol plan → Grok execute → Sol review","mean":4.845,"scored":6,"total":6,"same_family":true,"campaign_id":"v3-mixed-workflows-three-seed-20260713","reasoning_effort":"ultra","samples_per_cell":3},{"id":"sol-plan->terra-exec-sol-review","label":"Sol plan → Terra execute → Sol review","mean":5.133,"scored":6,"total":6,"same_family":true,"campaign_id":"v3-mixed-workflows-three-seed-20260713","reasoning_effort":"ultra","samples_per_cell":3},{"id":"sol-plan->terra-exec-grok-review","label":"Sol plan → Terra execute → Grok review","mean":4.567,"scored":6,"total":6,"same_family":true,"campaign_id":"v3-production-review-panels-three-seed-20260723","reasoning_effort":"ultra","samples_per_cell":3},{"id":"sol-plan@xhigh->exec@high+grok-review","label":"Sol plan → Sol execute → Sol + Grok review","mean":5.933,"scored":6,"total":6,"same_family":true,"campaign_id":"v3-production-review-panels-three-seed-20260723","reasoning_effort":"ultra","samples_per_cell":3},{"id":"all-deepseek-v4-pro-0813@xhigh","label":"DeepSeek V4 Pro 0813 xhigh","mean":3.239,"scored":6,"total":6,"same_family":false,"campaign_id":"v3-deepseek-v4-0813-20260813-r2","reasoning_effort":"ultra","samples_per_cell":3},{"id":"deepseek-pro@xhigh->flash@xhigh-exec+pro-review","label":"DeepSeek V4 Pro plan → V4 Flash execute → V4 Pro review","mean":3.339,"scored":6,"total":6,"same_family":false,"campaign_id":"v3-deepseek-v4-0813-20260813-r2","reasoning_effort":"ultra","samples_per_cell":3},{"id":"all-ox-alpha@high","label":"GLM 5.3 Flash (0x Alpha) via Pi high","mean":3.7945,"scored":6,"total":6,"same_family":false,"campaign_id":"v3-pi-ox-alpha-high-20260825-r3","reasoning_effort":"ultra","samples_per_cell":3},{"id":"all-ox-alpha@max","label":"GLM 5.3 Flash (0x Alpha) via Pi max","mean":3.5723,"scored":6,"total":6,"same_family":false,"campaign_id":"v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed","reasoning_effort":"ultra","samples_per_cell":3}],"reasoning_effort_by_campaign":{"v2-ce":"xhigh","v3-mixed-workflows-three-seed-20260713":"ultra","v3-production-review-panels-three-seed-20260723":"ultra","v3-deepseek-v4-0813-20260813-r2":"ultra","v3-pi-ox-alpha-high-20260825-r3":"ultra","v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed":"ultra"},"samples_per_cell_by_campaign":{"v2-ce":1,"v3-mixed-workflows-three-seed-20260713":3,"v3-production-review-panels-three-seed-20260723":3,"v3-deepseek-v4-0813-20260813-r2":3,"v3-pi-ox-alpha-high-20260825-r3":3,"v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed":3}},{"id":"fable-5","label":"Fable 5","reasoning_effort":"varies by campaign","samples_per_cell":null,"rows":[{"id":"all-codex-5.6-sol-xhigh","label":"GPT-5.6 Sol xhigh","mean":6.917,"scored":6,"total":6,"same_family":false,"campaign_id":"v2-ce","reasoning_effort":"xhigh","samples_per_cell":1},{"id":"all-grok-4.5","label":"Grok 4.5 xhigh","mean":6.333,"scored":6,"total":6,"same_family":false,"campaign_id":"v2-ce","reasoning_effort":"xhigh","samples_per_cell":1},{"id":"opus-plan->codex-exec-xhigh","label":"Opus plan → Codex 5.5 xhigh","mean":6.167,"scored":6,"total":6,"same_family":true,"campaign_id":"v2-ce","reasoning_effort":"xhigh","samples_per_cell":1},{"id":"all-opus-4.8","label":"Opus 4.8","mean":6.083,"scored":6,"total":6,"same_family":true,"campaign_id":"v2-ce","reasoning_effort":"xhigh","samples_per_cell":1},{"id":"all-glm-5.2","label":"GLM 5.2","mean":4.833,"scored":6,"total":6,"same_family":false,"campaign_id":"v2-ce","reasoning_effort":"xhigh","samples_per_cell":1},{"id":"all-codex-xhigh","label":"Codex 5.5 xhigh","mean":4.667,"scored":6,"total":6,"same_family":false,"campaign_id":"v2-ce","reasoning_effort":"xhigh","samples_per_cell":1},{"id":"fable-plan->grok-exec-sol-review","label":"Fable plan → Grok execute → Sol review","mean":6.194,"scored":6,"total":6,"same_family":true,"campaign_id":"v3-mixed-workflows-three-seed-20260713","reasoning_effort":"xhigh","samples_per_cell":3},{"id":"sol-plan->grok-exec-sol-review","label":"Sol plan → Grok execute → Sol review","mean":7.778,"scored":6,"total":6,"same_family":false,"campaign_id":"v3-mixed-workflows-three-seed-20260713","reasoning_effort":"xhigh","samples_per_cell":3},{"id":"sol-plan->terra-exec-sol-review","label":"Sol plan → Terra execute → Sol review","mean":7.055,"scored":6,"total":6,"same_family":false,"campaign_id":"v3-mixed-workflows-three-seed-20260713","reasoning_effort":"xhigh","samples_per_cell":3},{"id":"sol-plan->terra-exec-grok-review","label":"Sol plan → Terra execute → Grok review","mean":5.861,"scored":6,"total":6,"same_family":false,"campaign_id":"v3-production-review-panels-three-seed-20260723","reasoning_effort":"xhigh","samples_per_cell":3},{"id":"sol-plan@xhigh->exec@high+grok-review","label":"Sol plan → Sol execute → Sol + Grok review","mean":6.722,"scored":6,"total":6,"same_family":false,"campaign_id":"v3-production-review-panels-three-seed-20260723","reasoning_effort":"xhigh","samples_per_cell":3},{"id":"all-deepseek-v4-pro-0813@xhigh","label":"DeepSeek V4 Pro 0813 xhigh","mean":7.722,"scored":6,"total":6,"same_family":false,"campaign_id":"v3-deepseek-v4-0813-20260813-r2","reasoning_effort":"unspecified","samples_per_cell":3},{"id":"deepseek-pro@xhigh->flash@xhigh-exec+pro-review","label":"DeepSeek V4 Pro plan → V4 Flash execute → V4 Pro review","mean":6.333,"scored":6,"total":6,"same_family":false,"campaign_id":"v3-deepseek-v4-0813-20260813-r2","reasoning_effort":"unspecified","samples_per_cell":3},{"id":"all-ox-alpha@high","label":"GLM 5.3 Flash (0x Alpha) via Pi high","mean":6.3055,"scored":6,"total":6,"same_family":false,"campaign_id":"v3-pi-ox-alpha-high-20260825-r3","reasoning_effort":"unspecified","samples_per_cell":3},{"id":"all-ox-alpha@max","label":"GLM 5.3 Flash (0x Alpha) via Pi max","mean":6.5,"scored":6,"total":6,"same_family":false,"campaign_id":"v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed","reasoning_effort":"unspecified","samples_per_cell":3}],"reasoning_effort_by_campaign":{"v2-ce":"xhigh","v3-mixed-workflows-three-seed-20260713":"xhigh","v3-production-review-panels-three-seed-20260723":"xhigh","v3-deepseek-v4-0813-20260813-r2":"unspecified","v3-pi-ox-alpha-high-20260825-r3":"unspecified","v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed":"unspecified"},"samples_per_cell_by_campaign":{"v2-ce":1,"v3-mixed-workflows-three-seed-20260713":3,"v3-production-review-panels-three-seed-20260723":3,"v3-deepseek-v4-0813-20260813-r2":3,"v3-pi-ox-alpha-high-20260825-r3":3,"v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed":3}}],"supplemental_judge":{"id":"gpt-5.5-pro","label":"GPT-5.5 Pro","note":"Historical supplemental ruler; it scored only the add-i-key cell for Codex 5.5 xhigh, GLM 5.2, and Grok 4.5 and is not used to rank the full slate.","rows":[{"id":"all-codex-xhigh","label":"Codex 5.5 xhigh","mean":5,"scored":1,"total":6,"same_family":true},{"id":"all-glm-5.2","label":"GLM 5.2","mean":4,"scored":1,"total":6,"same_family":false},{"id":"all-grok-4.5","label":"Grok 4.5 xhigh","mean":4,"scored":1,"total":6,"same_family":false}]},"score_order":["fable-5","gpt-5.6-sol"],"tasks":[{"key":"add-i-key","id":"add-i-key-with-legend-260522-ca28","type":"feature","desc":"Add an i key and legend entry that open a full task-information panel in the Hive TUI.","pr":"https://github.com/ivankuznetsov/hive/pull/127","plan_author":"claude","original_model":"claude-opus-4-7"},{"key":"web-install","id":"add-local-hive-web-install-260629-f4ca","type":"feature","desc":"Add a first-class local, non-Docker install and run mode for the Hive web UI.","pr":"https://github.com/ivankuznetsov/hive/pull/622","plan_author":"claude","original_model":"gpt-5.5 (Codex)"},{"key":"install","id":"figure-out-way-to-install-260513-4a0a","type":"feature","desc":"Package Hive for straightforward installation on macOS and several Linux distributions.","pr":"https://github.com/ivankuznetsov/hive/pull/103","plan_author":"claude","original_model":"claude-opus-4-7"},{"key":"fix-tmux","id":"fix-claude-tmux-ready-detector-260629-50cc","type":"bugfix","desc":"Fix missing launcher scripts and ready-prompt detection for Claude Code in tmux mode.","pr":"https://github.com/ivankuznetsov/hive/pull/623","plan_author":"claude","original_model":"gpt-5.5 (Codex)"},{"key":"fix-review","id":"fix-review-stage-claude-stop-260629-26ed","type":"bugfix","desc":"Recover completed review passes from Claude stop-hook failures instead of leaving REVIEW_ERROR.","pr":"https://github.com/ivankuznetsov/hive/pull/625","plan_author":"claude","original_model":"gpt-5.5 (Codex)"},{"key":"daemon","id":"make-the-hive-daemon-automatically-260629-223d","type":"feature","desc":"Make the Hive daemon retry recoverable terminal errors after the failing dependency becomes healthy.","pr":"https://github.com/ivankuznetsov/hive/pull/624","plan_author":"claude","original_model":"gpt-5.5 (Codex)"}],"candidates":[{"id":"all-codex-5.6-sol-xhigh","label":"GPT-5.6 Sol xhigh","reasoning":"xhigh (pinned)","note":"Sol drives plan, execute, and review.","completed":6,"total":6,"mean_minutes":146.9,"time_sample":4,"cost_per_task_usd":47.66,"cost_total_usd":285.9714,"cost_sample":6,"normalized_mtokens_per_task":74.86,"token_sample":6,"token_totals":{"input":7128536,"output":993973,"cache_read":441019136,"cache_write":0},"efficiency_by_task":{"add-i-key":{"cost_usd":20.9874,"normalized_mtokens":28.26,"tokens":{"input":735389,"output":120344,"cache_read":27400192,"cache_write":0},"wall_minutes":95.3},"web-install":{"cost_usd":88.2255,"normalized_mtokens":142.43,"tokens":{"input":2161848,"output":246859,"cache_read":140020992,"cache_write":0},"wall_minutes":227.4},"install":{"cost_usd":105.2448,"normalized_mtokens":167.52,"tokens":{"input":2388996,"output":363868,"cache_read":164767488,"cache_write":0},"wall_minutes":224.2},"fix-tmux":{"cost_usd":20.3585,"normalized_mtokens":32.37,"tokens":{"input":580766,"output":52902,"cache_read":31735296,"cache_write":0},"wall_minutes":40.7},"fix-review":{"cost_usd":18.9374,"normalized_mtokens":29.98,"tokens":{"input":437504,"output":67028,"cache_read":29478144,"cache_write":0},"wall_minutes":null},"daemon":{"cost_usd":32.2178,"normalized_mtokens":48.58,"tokens":{"input":824033,"output":142972,"cache_read":47617024,"cache_write":0},"wall_minutes":null}},"cells":{"add-i-key":{"scores":"7.0 / 7.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-codex-5.6-sol-xhigh--add-i-key-with-legend-260522-ca28/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-codex-5.6-sol-xhigh--add-i-key-with-legend-260522-ca28/candidate.patch"},"web-install":{"scores":"7.0 / 5.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-codex-5.6-sol-xhigh--add-local-hive-web-install-260629-f4ca/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-codex-5.6-sol-xhigh--add-local-hive-web-install-260629-f4ca/candidate.patch"},"install":{"scores":"5.5 / 3.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-codex-5.6-sol-xhigh--figure-out-way-to-install-260513-4a0a/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-codex-5.6-sol-xhigh--figure-out-way-to-install-260513-4a0a/candidate.patch"},"fix-tmux":{"scores":"9.0 / 8.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-codex-5.6-sol-xhigh--fix-claude-tmux-ready-detector-260629-50cc/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-codex-5.6-sol-xhigh--fix-claude-tmux-ready-detector-260629-50cc/candidate.patch"},"fix-review":{"scores":"6.0 / 7.1","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-codex-5.6-sol-xhigh--fix-review-stage-claude-stop-260629-26ed/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-codex-5.6-sol-xhigh--fix-review-stage-claude-stop-260629-26ed/candidate.patch"},"daemon":{"scores":"7.0 / 5.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-codex-5.6-sol-xhigh--make-the-hive-daemon-automatically-260629-223d/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-codex-5.6-sol-xhigh--make-the-hive-daemon-automatically-260629-223d/candidate.patch"}},"campaign_id":"v2-ce","samples_per_cell":1,"raw_evidence_published":true},{"id":"opus-plan->codex-exec-xhigh","label":"Opus plan → Codex 5.5 xhigh","reasoning":"Opus default · Codex xhigh","note":"Opus plans and reviews; Codex 5.5 xhigh executes. Cost is attributed per model from all six stage-log sets.","completed":6,"total":6,"mean_minutes":60.8,"time_sample":5,"cost_per_task_usd":19.26,"cost_total_usd":115.5829,"cost_sample":6,"normalized_mtokens_per_task":27.67,"token_sample":6,"token_totals":{"input":2316931,"output":328495,"cache_read":160063061,"cache_write":3323418},"cost_breakdown_usd":{"gpt-5.5":67.3351,"claude-opus-4-8":47.6319,"claude-haiku-4-5":0.6159},"efficiency_by_task":{"add-i-key":{"cost_usd":25.7455,"normalized_mtokens":39.89,"tokens":{"input":392117,"output":50805,"cache_read":38684505,"cache_write":759794},"wall_minutes":null},"web-install":{"cost_usd":16.9525,"normalized_mtokens":22.79,"tokens":{"input":480800,"output":86830,"cache_read":22032050,"cache_write":186130},"wall_minutes":50.9},"install":{"cost_usd":31.017,"normalized_mtokens":47.85,"tokens":{"input":360883,"output":59578,"cache_read":46605064,"cache_write":827662},"wall_minutes":75},"fix-tmux":{"cost_usd":11.9163,"normalized_mtokens":14.69,"tokens":{"input":253051,"output":26094,"cache_read":13814139,"cache_write":593882},"wall_minutes":73},"fix-review":{"cost_usd":12.948,"normalized_mtokens":18.12,"tokens":{"input":315214,"output":42512,"cache_read":17486693,"cache_write":270877},"wall_minutes":46.9},"daemon":{"cost_usd":17.0037,"normalized_mtokens":22.7,"tokens":{"input":514866,"output":62676,"cache_read":21440610,"cache_write":685073},"wall_minutes":58.2}},"cells":{"add-i-key":{"scores":"6.0 / 5.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/opus-plan-%3Ecodex-exec-xhigh--add-i-key-with-legend-260522-ca28/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/opus-plan-%3Ecodex-exec-xhigh--add-i-key-with-legend-260522-ca28/candidate.patch"},"web-install":{"scores":"5.0 / 3.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/opus-plan-%3Ecodex-exec-xhigh--add-local-hive-web-install-260629-f4ca/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/opus-plan-%3Ecodex-exec-xhigh--add-local-hive-web-install-260629-f4ca/candidate.patch"},"install":{"scores":"6.5 / 4.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/opus-plan-%3Ecodex-exec-xhigh--figure-out-way-to-install-260513-4a0a/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/opus-plan-%3Ecodex-exec-xhigh--figure-out-way-to-install-260513-4a0a/candidate.patch"},"fix-tmux":{"scores":"9.0 / 8.7","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/opus-plan-%3Ecodex-exec-xhigh--fix-claude-tmux-ready-detector-260629-50cc/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/opus-plan-%3Ecodex-exec-xhigh--fix-claude-tmux-ready-detector-260629-50cc/candidate.patch"},"fix-review":{"scores":"4.0 / 3.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/opus-plan-%3Ecodex-exec-xhigh--fix-review-stage-claude-stop-260629-26ed/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/opus-plan-%3Ecodex-exec-xhigh--fix-review-stage-claude-stop-260629-26ed/candidate.patch"},"daemon":{"scores":"6.5 / 4.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/opus-plan-%3Ecodex-exec-xhigh--make-the-hive-daemon-automatically-260629-223d/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/opus-plan-%3Ecodex-exec-xhigh--make-the-hive-daemon-automatically-260629-223d/candidate.patch"}},"campaign_id":"v2-ce","samples_per_cell":1,"raw_evidence_published":true},{"id":"all-opus-4.8","label":"Opus 4.8","reasoning":"Claude CLI default","note":"Opus drives plan, execute, and review.","completed":6,"total":6,"mean_minutes":63.7,"time_sample":6,"cost_per_task_usd":35.74,"cost_total_usd":214.4371,"cost_sample":6,"normalized_mtokens_per_task":59.82,"token_sample":6,"token_totals":{"input":923038,"output":36000,"cache_read":349561272,"cache_write":8424071},"efficiency_by_task":{"add-i-key":{"cost_usd":24.2902,"normalized_mtokens":39.03,"tokens":{"input":139286,"output":5040,"cache_read":37692192,"cache_write":1194544},"wall_minutes":59.5},"web-install":{"cost_usd":33.0767,"normalized_mtokens":54.71,"tokens":{"input":117334,"output":5300,"cache_read":53063280,"cache_write":1523627},"wall_minutes":44},"install":{"cost_usd":74.3344,"normalized_mtokens":125.82,"tokens":{"input":279090,"output":10667,"cache_read":123327646,"cache_write":2201693},"wall_minutes":98.7},"fix-tmux":{"cost_usd":11.3479,"normalized_mtokens":17.23,"tokens":{"input":128227,"output":2492,"cache_read":16384230,"cache_write":714939},"wall_minutes":39.9},"fix-review":{"cost_usd":27.7496,"normalized_mtokens":47.07,"tokens":{"input":116038,"output":5343,"cache_read":45933297,"cache_write":1017808},"wall_minutes":55.1},"daemon":{"cost_usd":43.6381,"normalized_mtokens":75.08,"tokens":{"input":143063,"output":7158,"cache_read":73160627,"cache_write":1771460},"wall_minutes":85}},"cells":{"add-i-key":{"scores":"6.0 / 5.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-opus-4.8--add-i-key-with-legend-260522-ca28/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-opus-4.8--add-i-key-with-legend-260522-ca28/candidate.patch"},"web-install":{"scores":"4.5 / 2.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-opus-4.8--add-local-hive-web-install-260629-f4ca/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-opus-4.8--add-local-hive-web-install-260629-f4ca/candidate.patch"},"install":{"scores":"6.0 / 2.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-opus-4.8--figure-out-way-to-install-260513-4a0a/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-opus-4.8--figure-out-way-to-install-260513-4a0a/candidate.patch"},"fix-tmux":{"scores":"8.0 / 7.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-opus-4.8--fix-claude-tmux-ready-detector-260629-50cc/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-opus-4.8--fix-claude-tmux-ready-detector-260629-50cc/candidate.patch"},"fix-review":{"scores":"5.0 / 3.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-opus-4.8--fix-review-stage-claude-stop-260629-26ed/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-opus-4.8--fix-review-stage-claude-stop-260629-26ed/candidate.patch"},"daemon":{"scores":"7.0 / 5.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-opus-4.8--make-the-hive-daemon-automatically-260629-223d/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-opus-4.8--make-the-hive-daemon-automatically-260629-223d/candidate.patch"}},"campaign_id":"v2-ce","samples_per_cell":1,"raw_evidence_published":true},{"id":"all-grok-4.5","label":"Grok 4.5 xhigh","reasoning":"xhigh (pinned)","note":"Grok drives plan, execute, and review; usage tokens and cost are unavailable.","completed":6,"total":6,"mean_minutes":27.3,"time_sample":5,"cost_per_task_usd":null,"cost_total_usd":null,"cost_sample":0,"normalized_mtokens_per_task":null,"token_sample":0,"token_totals":null,"efficiency_by_task":{"add-i-key":{"cost_usd":null,"normalized_mtokens":null,"tokens":null,"wall_minutes":30.1},"web-install":{"cost_usd":null,"normalized_mtokens":null,"tokens":null,"wall_minutes":32.1},"install":{"cost_usd":null,"normalized_mtokens":null,"tokens":null,"wall_minutes":27.3},"fix-tmux":{"cost_usd":null,"normalized_mtokens":null,"tokens":null,"wall_minutes":null},"fix-review":{"cost_usd":null,"normalized_mtokens":null,"tokens":null,"wall_minutes":23.2},"daemon":{"cost_usd":null,"normalized_mtokens":null,"tokens":null,"wall_minutes":23.9}},"cells":{"add-i-key":{"scores":"5.5 / 5.1","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-grok-4.5--add-i-key-with-legend-260522-ca28/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-grok-4.5--add-i-key-with-legend-260522-ca28/candidate.patch"},"web-install":{"scores":"6.0 / 4.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-grok-4.5--add-local-hive-web-install-260629-f4ca/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-grok-4.5--add-local-hive-web-install-260629-f4ca/candidate.patch"},"install":{"scores":"6.5 / 2.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-grok-4.5--figure-out-way-to-install-260513-4a0a/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-grok-4.5--figure-out-way-to-install-260513-4a0a/candidate.patch"},"fix-tmux":{"scores":"7.5 / 6.2","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-grok-4.5--fix-claude-tmux-ready-detector-260629-50cc/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-grok-4.5--fix-claude-tmux-ready-detector-260629-50cc/candidate.patch"},"fix-review":{"scores":"5.0 / 3.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-grok-4.5--fix-review-stage-claude-stop-260629-26ed/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-grok-4.5--fix-review-stage-claude-stop-260629-26ed/candidate.patch"},"daemon":{"scores":"7.5 / 5.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-grok-4.5--make-the-hive-daemon-automatically-260629-223d/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-grok-4.5--make-the-hive-daemon-automatically-260629-223d/candidate.patch"}},"campaign_id":"v2-ce","samples_per_cell":1,"raw_evidence_published":true},{"id":"all-codex-xhigh","label":"Codex 5.5 xhigh","reasoning":"xhigh (pinned)","note":"GPT-5.5 drives plan, execute, and review.","completed":6,"total":6,"mean_minutes":52,"time_sample":6,"cost_per_task_usd":17.17,"cost_total_usd":102.9917,"cost_sample":6,"normalized_mtokens_per_task":24.01,"token_sample":6,"token_totals":{"input":3342536,"output":539602,"cache_read":140181888,"cache_write":0},"efficiency_by_task":{"add-i-key":{"cost_usd":14.5531,"normalized_mtokens":19.17,"tokens":{"input":563535,"output":82433,"cache_read":18524800,"cache_write":0},"wall_minutes":51},"web-install":{"cost_usd":20.9505,"normalized_mtokens":29.53,"tokens":{"input":709534,"output":101380,"cache_read":28722816,"cache_write":0},"wall_minutes":52.9},"install":{"cost_usd":20.2089,"normalized_mtokens":28.59,"tokens":{"input":547432,"output":116995,"cache_read":27923712,"cache_write":0},"wall_minutes":67},"fix-tmux":{"cost_usd":7.8062,"normalized_mtokens":11.42,"tokens":{"input":229784,"output":35994,"cache_read":11154944,"cache_write":0},"wall_minutes":26},"fix-review":{"cost_usd":22.6186,"normalized_mtokens":31.1,"tokens":{"input":795250,"output":118271,"cache_read":30188416,"cache_write":0},"wall_minutes":75},"daemon":{"cost_usd":16.8545,"normalized_mtokens":24.25,"tokens":{"input":497001,"output":84529,"cache_read":23667200,"cache_write":0},"wall_minutes":40.4}},"cells":{"add-i-key":{"scores":"5.5 / 4.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-codex-xhigh--add-i-key-with-legend-260522-ca28/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-codex-xhigh--add-i-key-with-legend-260522-ca28/candidate.patch"},"web-install":{"scores":"3.0 / 3.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-codex-xhigh--add-local-hive-web-install-260629-f4ca/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-codex-xhigh--add-local-hive-web-install-260629-f4ca/candidate.patch"},"install":{"scores":"5.0 / 2.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-codex-xhigh--figure-out-way-to-install-260513-4a0a/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-codex-xhigh--figure-out-way-to-install-260513-4a0a/candidate.patch"},"fix-tmux":{"scores":"5.0 / 6.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-codex-xhigh--fix-claude-tmux-ready-detector-260629-50cc/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-codex-xhigh--fix-claude-tmux-ready-detector-260629-50cc/candidate.patch"},"fix-review":{"scores":"3.5 / 2.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-codex-xhigh--fix-review-stage-claude-stop-260629-26ed/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-codex-xhigh--fix-review-stage-claude-stop-260629-26ed/candidate.patch"},"daemon":{"scores":"6.0 / 4.2","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-codex-xhigh--make-the-hive-daemon-automatically-260629-223d/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-codex-xhigh--make-the-hive-daemon-automatically-260629-223d/candidate.patch"}},"campaign_id":"v2-ce","samples_per_cell":1,"raw_evidence_published":true},{"id":"all-glm-5.2","label":"GLM 5.2","reasoning":"provider default","note":"GLM drives plan, execute, and review through pi.","completed":6,"total":6,"mean_minutes":114.4,"time_sample":6,"cost_per_task_usd":5.65,"cost_total_usd":33.9223,"cost_sample":6,"normalized_mtokens_per_task":21.793,"token_sample":6,"token_totals":{"input":10874917,"output":713673,"cache_read":119167854,"cache_write":0},"efficiency_by_task":{"add-i-key":{"cost_usd":4.2748,"normalized_mtokens":17.548,"tokens":{"input":1111943,"output":92230,"cache_read":16343368,"cache_write":0},"wall_minutes":114.1},"web-install":{"cost_usd":5.5608,"normalized_mtokens":24.145,"tokens":{"input":1090293,"output":133044,"cache_read":22921488,"cache_write":0},"wall_minutes":173.5},"install":{"cost_usd":6.8015,"normalized_mtokens":27.155,"tokens":{"input":1998882,"output":132781,"cache_read":25023534,"cache_write":0},"wall_minutes":124.4},"fix-tmux":{"cost_usd":2.4695,"normalized_mtokens":10.484,"tokens":{"input":351009,"output":110691,"cache_read":10022251,"cache_write":0},"wall_minutes":91.4},"fix-review":{"cost_usd":3.1289,"normalized_mtokens":4.316,"tokens":{"input":2929630,"output":34099,"cache_read":1352715,"cache_write":0},"wall_minutes":29.7},"daemon":{"cost_usd":11.6868,"normalized_mtokens":47.108,"tokens":{"input":3393160,"output":210828,"cache_read":43504498,"cache_write":0},"wall_minutes":153.4}},"cells":{"add-i-key":{"scores":"6.5 / 5.5","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-glm-5.2--add-i-key-with-legend-260522-ca28/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-glm-5.2--add-i-key-with-legend-260522-ca28/candidate.patch"},"web-install":{"scores":"3.0 / 3.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-glm-5.2--add-local-hive-web-install-260629-f4ca/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-glm-5.2--add-local-hive-web-install-260629-f4ca/candidate.patch"},"install":{"scores":"6.0 / 2.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-glm-5.2--figure-out-way-to-install-260513-4a0a/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-glm-5.2--figure-out-way-to-install-260513-4a0a/candidate.patch"},"fix-tmux":{"scores":"7.0 / 7.2","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-glm-5.2--fix-claude-tmux-ready-detector-260629-50cc/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-glm-5.2--fix-claude-tmux-ready-detector-260629-50cc/candidate.patch"},"fix-review":{"scores":"1.5 / 1.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-glm-5.2--fix-review-stage-claude-stop-260629-26ed/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-glm-5.2--fix-review-stage-claude-stop-260629-26ed/candidate.patch"},"daemon":{"scores":"5.0 / 3.0","result_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/all-glm-5.2--make-the-hive-daemon-automatically-260629-223d/results.json","patch_url":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/cells/all-glm-5.2--make-the-hive-daemon-automatically-260629-223d/candidate.patch"}},"campaign_id":"v2-ce","samples_per_cell":1,"raw_evidence_published":true},{"id":"fable-plan->grok-exec-sol-review","label":"Fable plan → Grok execute → Sol review","reasoning":"Fable high plan · Grok xhigh execute · Sol xhigh review","note":"Fable plans, Grok executes, and Sol performs the production review. Known-provider token and cost subtotals include Fable planning plus retained Sol review events; Grok provides no usable usage telemetry.","token_scope":"Fable plan + Sol review only · Grok telemetry unavailable","campaign_id":"v3-mixed-workflows-three-seed-20260713","samples_per_cell":3,"raw_evidence_published":false,"completed":6,"total":6,"mean_minutes":98.4,"time_sample":5,"cost_per_task_usd":null,"cost_total_usd":null,"cost_sample":0,"normalized_mtokens_per_task":9.24,"token_sample":6,"token_totals":{"input":960135,"output":203565,"cache_read":52219710,"cache_write":2059137},"efficiency_by_task":{"add-i-key":{"cost_usd":null,"normalized_mtokens":14.114,"tokens":{"input":341374,"output":47871,"cache_read":13463399,"cache_write":260906},"wall_minutes":92.1,"known_cost_usd":11.1775},"web-install":{"cost_usd":null,"normalized_mtokens":5.927,"tokens":{"input":23367,"output":1834,"cache_read":5250040,"cache_write":651537},"wall_minutes":150.7,"known_cost_usd":3.4452},"install":{"cost_usd":null,"normalized_mtokens":1.335,"tokens":{"input":14945,"output":70864,"cache_read":1046597,"cache_write":202794},"wall_minutes":null,"known_cost_usd":3.3836},"fix-tmux":{"cost_usd":null,"normalized_mtokens":2.359,"tokens":{"input":30336,"output":290,"cache_read":2083183,"cache_write":245466},"wall_minutes":25.5,"known_cost_usd":2.4279},"fix-review":{"cost_usd":null,"normalized_mtokens":28.061,"tokens":{"input":524767,"output":81985,"cache_read":27130738,"cache_write":323278},"wall_minutes":87.2,"known_cost_usd":20.2622},"daemon":{"cost_usd":null,"normalized_mtokens":3.647,"tokens":{"input":25346,"output":721,"cache_read":3245753,"cache_write":375156},"wall_minutes":136.2,"known_cost_usd":1.9333}},"cells":{"add-i-key":{"scores":"8.5 / 6.0","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/fable-plan-grok-exec-sol-review/add-i-key.patch","judge_samples":{"fable-5":{"mean":8.5,"stddev":0,"scores":[8.5,8.5,8.5],"interval":[8.5,8.5],"same_family":true,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":6,"stddev":0.408,"scores":[6,6.5,5.5],"interval":[5.592,6.408],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"web-install":{"scores":"5.167 / 3.167","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/fable-plan-grok-exec-sol-review/web-install.patch","judge_samples":{"fable-5":{"mean":5.167,"stddev":0.85,"scores":[4,6,5.5],"interval":[4.317,6.017],"same_family":true,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":3.167,"stddev":0.236,"scores":[3,3.5,3],"interval":[2.931,3.402],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"install":{"scores":"5.5 / 2.833","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/fable-plan-grok-exec-sol-review/install.patch","judge_samples":{"fable-5":{"mean":5.5,"stddev":0.408,"scores":[5,6,5.5],"interval":[5.092,5.908],"same_family":true,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":2.833,"stddev":0.236,"scores":[3,3,2.5],"interval":[2.598,3.069],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":true},"fix-tmux":{"scores":"5.833 / 7.833","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/fable-plan-grok-exec-sol-review/fix-tmux.patch","judge_samples":{"fable-5":{"mean":5.833,"stddev":0.236,"scores":[5.5,6,6],"interval":[5.598,6.069],"same_family":true,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":7.833,"stddev":0.236,"scores":[7.5,8,8],"interval":[7.598,8.069],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"fix-review":{"scores":"5.833 / 4.0","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/fable-plan-grok-exec-sol-review/fix-review.patch","judge_samples":{"fable-5":{"mean":5.833,"stddev":0.236,"scores":[5.5,6,6],"interval":[5.598,6.069],"same_family":true,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":4,"stddev":0.408,"scores":[4.5,4,3.5],"interval":[3.592,4.408],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"daemon":{"scores":"6.333 / 4.167","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/fable-plan-grok-exec-sol-review/daemon.patch","judge_samples":{"fable-5":{"mean":6.333,"stddev":0.236,"scores":[6.5,6.5,6],"interval":[6.098,6.569],"same_family":true,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":4.167,"stddev":0.236,"scores":[4,4,4.5],"interval":[3.931,4.402],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false}},"patches_published":true,"known_cost_per_task_usd":7.10495,"known_cost_total_usd":42.6297,"known_cost_sample":6},{"id":"sol-plan->grok-exec-sol-review","label":"Sol plan → Grok execute → Sol review","reasoning":"Sol xhigh plan · Grok xhigh execute · Sol xhigh review","note":"Sol plans and reviews while Grok executes. Token figures cover Sol stages only; Grok provides no usable usage telemetry.","token_scope":"Sol plan/review only · Grok telemetry unavailable","campaign_id":"v3-mixed-workflows-three-seed-20260713","samples_per_cell":3,"raw_evidence_published":false,"completed":6,"total":6,"mean_minutes":132.1,"time_sample":5,"cost_per_task_usd":null,"cost_total_usd":null,"cost_sample":0,"normalized_mtokens_per_task":29.123,"token_sample":6,"token_totals":{"input":4903724,"output":633331,"cache_read":169201408,"cache_write":0},"efficiency_by_task":{"add-i-key":{"cost_usd":null,"normalized_mtokens":33.998,"tokens":{"input":802017,"output":116378,"cache_read":33079296,"cache_write":0},"wall_minutes":115.5,"known_cost_usd":24.0411},"web-install":{"cost_usd":null,"normalized_mtokens":29.397,"tokens":{"input":1077755,"output":91952,"cache_read":28227584,"cache_write":0},"wall_minutes":94.9,"known_cost_usd":22.2611},"install":{"cost_usd":null,"normalized_mtokens":30.182,"tokens":{"input":1031969,"output":153008,"cache_read":28996608,"cache_write":0},"wall_minutes":173.5,"known_cost_usd":24.2484},"fix-tmux":{"cost_usd":null,"normalized_mtokens":6.697,"tokens":{"input":281295,"output":32892,"cache_read":6382848,"cache_write":0},"wall_minutes":63.5,"known_cost_usd":5.5847},"fix-review":{"cost_usd":null,"normalized_mtokens":22.795,"tokens":{"input":502659,"output":84478,"cache_read":22207744,"cache_write":0},"wall_minutes":null,"known_cost_usd":16.1515},"daemon":{"cost_usd":null,"normalized_mtokens":51.67,"tokens":{"input":1208029,"output":154623,"cache_read":50307328,"cache_write":0},"wall_minutes":213.1,"known_cost_usd":35.8325}},"cells":{"add-i-key":{"scores":"9.0 / 6.167","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/sol-plan-grok-exec-sol-review/add-i-key.patch","judge_samples":{"fable-5":{"mean":9,"stddev":0,"scores":[9,9,9],"interval":[9,9],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":6.167,"stddev":0.236,"scores":[6.5,6,6],"interval":[5.931,6.402],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"web-install":{"scores":"8.333 / 5.167","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/sol-plan-grok-exec-sol-review/web-install.patch","judge_samples":{"fable-5":{"mean":8.333,"stddev":0.236,"scores":[8,8.5,8.5],"interval":[8.098,8.569],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":5.167,"stddev":0.236,"scores":[5,5.5,5],"interval":[4.931,5.402],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"install":{"scores":"7.833 / 4.067","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/sol-plan-grok-exec-sol-review/install.patch","judge_samples":{"fable-5":{"mean":7.833,"stddev":0.471,"scores":[7.5,7.5,8.5],"interval":[7.362,8.305],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":4.067,"stddev":0.094,"scores":[4,4.2,4],"interval":[3.972,4.161],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"fix-tmux":{"scores":"9.167 / 5.833","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/sol-plan-grok-exec-sol-review/fix-tmux.patch","judge_samples":{"fable-5":{"mean":9.167,"stddev":0.236,"scores":[9,9.5,9],"interval":[8.931,9.402],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":5.833,"stddev":0.236,"scores":[5.5,6,6],"interval":[5.598,6.069],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"fix-review":{"scores":"4.0 / 4.0","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/sol-plan-grok-exec-sol-review/fix-review.patch","judge_samples":{"fable-5":{"mean":4,"stddev":0.408,"scores":[4.5,3.5,4],"interval":[3.592,4.408],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":4,"stddev":0,"scores":[4,4,4],"interval":[4,4],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":true},"daemon":{"scores":"8.333 / 3.833","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/sol-plan-grok-exec-sol-review/daemon.patch","judge_samples":{"fable-5":{"mean":8.333,"stddev":0.236,"scores":[8,8.5,8.5],"interval":[8.098,8.569],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":3.833,"stddev":0.85,"scores":[5,3.5,3],"interval":[2.983,4.683],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false}},"patches_published":true,"known_cost_per_task_usd":21.353217,"known_cost_total_usd":128.1193,"known_cost_sample":6},{"id":"sol-plan->terra-exec-sol-review","label":"Sol plan → Terra execute → Sol review","reasoning":"Sol xhigh plan · Terra xhigh execute · Sol xhigh review","note":"Sol plans and reviews while Terra executes.","campaign_id":"v3-mixed-workflows-three-seed-20260713","samples_per_cell":3,"raw_evidence_published":false,"completed":6,"total":6,"mean_minutes":166,"time_sample":4,"cost_per_task_usd":26.12,"cost_total_usd":130.597,"cost_sample":5,"normalized_mtokens_per_task":34.807,"token_sample":5,"token_totals":{"input":4947853,"output":722519,"cache_read":168364288,"cache_write":0},"efficiency_by_task":{"add-i-key":{"cost_usd":17.1596,"normalized_mtokens":21.823,"tokens":{"input":688562,"output":106769,"cache_read":21027328,"cache_write":0},"wall_minutes":91.4},"web-install":{"cost_usd":21.697,"normalized_mtokens":30.491,"tokens":{"input":748275,"output":104556,"cache_read":29637888,"cache_write":0},"wall_minutes":197.7},"install":{"cost_usd":31.5074,"normalized_mtokens":38.249,"tokens":{"input":1459063,"output":197194,"cache_read":36592640,"cache_write":0},"wall_minutes":179.5},"fix-tmux":{"cost_usd":31.2778,"normalized_mtokens":43.864,"tokens":{"input":1031246,"output":159491,"cache_read":42673664,"cache_write":0},"wall_minutes":null},"fix-review":{"cost_usd":null,"normalized_mtokens":null,"tokens":null,"wall_minutes":null},"daemon":{"cost_usd":28.9552,"normalized_mtokens":39.608,"tokens":{"input":1020707,"output":154509,"cache_read":38432768,"cache_write":0},"wall_minutes":195.2}},"cells":{"add-i-key":{"scores":"8.833 / 5.267","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/sol-plan-terra-exec-sol-review/add-i-key.patch","judge_samples":{"fable-5":{"mean":8.833,"stddev":0.236,"scores":[9,8.5,9],"interval":[8.598,9.069],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":5.267,"stddev":0.377,"scores":[5.8,5,5],"interval":[4.89,5.644],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"web-install":{"scores":"3.667 / 2.833","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/sol-plan-terra-exec-sol-review/web-install.patch","judge_samples":{"fable-5":{"mean":3.667,"stddev":0.471,"scores":[4,3,4],"interval":[3.195,4.138],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":2.833,"stddev":0.236,"scores":[2.5,3,3],"interval":[2.598,3.069],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"install":{"scores":"8.0 / 3.167","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/sol-plan-terra-exec-sol-review/install.patch","judge_samples":{"fable-5":{"mean":8,"stddev":0,"scores":[8,8,8],"interval":[8,8],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":3.167,"stddev":0.236,"scores":[3,3.5,3],"interval":[2.931,3.402],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"fix-tmux":{"scores":"8.333 / 6.9","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/sol-plan-terra-exec-sol-review/fix-tmux.patch","judge_samples":{"fable-5":{"mean":8.333,"stddev":0.236,"scores":[8.5,8.5,8],"interval":[8.098,8.569],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":6.9,"stddev":0.294,"scores":[6.5,7,7.2],"interval":[6.606,7.194],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":true},"fix-review":{"scores":"8.167 / 7.233","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/sol-plan-terra-exec-sol-review/fix-review.patch","judge_samples":{"fable-5":{"mean":8.167,"stddev":0.624,"scores":[7.5,8,9],"interval":[7.543,8.79],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":7.233,"stddev":0.613,"scores":[8,6.5,7.2],"interval":[6.621,7.846],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":true},"daemon":{"scores":"5.333 / 5.4","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-mixed-workflows-three-seed-20260713/sol-plan-terra-exec-sol-review/daemon.patch","judge_samples":{"fable-5":{"mean":5.333,"stddev":0.471,"scores":[6,5,5],"interval":[4.862,5.805],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":5.4,"stddev":0.698,"scores":[4.5,5.5,6.2],"interval":[4.702,6.098],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false}},"patches_published":true},{"id":"sol-plan->terra-exec-grok-review","label":"Sol plan → Terra execute → Grok review","reasoning":"Sol xhigh plan · Terra xhigh execute · Grok xhigh review","note":"Sol plans, Terra executes, and Grok owns open-PR, review, triage, and fixes.","campaign_id":"v3-production-review-panels-three-seed-20260723","samples_per_cell":3,"raw_evidence_published":false,"patches_published":true,"completed":6,"total":6,"mean_minutes":61.1,"time_sample":6,"cost_per_task_usd":null,"cost_total_usd":null,"cost_sample":0,"known_cost_per_task_usd":18.633017,"known_cost_total_usd":111.7981,"known_cost_sample":6,"normalized_mtokens_per_task":26.229,"token_sample":6,"token_scope":"Sol plan + Terra execute only · Grok review telemetry unavailable","token_totals":{"input":3550841,"output":580785,"cache_read":153240832,"cache_write":0},"efficiency_by_task":{"add-i-key":{"cost_usd":null,"known_cost_usd":12.5279,"normalized_mtokens":17.19,"tokens":{"input":471572,"output":61386,"cache_read":16656896,"cache_write":0},"wall_minutes":43.7},"web-install":{"cost_usd":null,"known_cost_usd":31.7767,"normalized_mtokens":48.013,"tokens":{"input":881946,"output":128867,"cache_read":47001856,"cache_write":0},"wall_minutes":75.1},"install":{"cost_usd":null,"known_cost_usd":20.609,"normalized_mtokens":28.593,"tokens":{"input":512088,"output":135865,"cache_read":27945216,"cache_write":0},"wall_minutes":54.3},"fix-tmux":{"cost_usd":null,"known_cost_usd":12.3028,"normalized_mtokens":17.293,"tokens":{"input":384691,"output":65263,"cache_read":16843008,"cache_write":0},"wall_minutes":93.4},"fix-review":{"cost_usd":null,"known_cost_usd":15.1567,"normalized_mtokens":20.982,"tokens":{"input":520195,"output":78813,"cache_read":20382720,"cache_write":0},"wall_minutes":50.9},"daemon":{"cost_usd":null,"known_cost_usd":19.425,"normalized_mtokens":25.302,"tokens":{"input":780349,"output":110591,"cache_read":24411136,"cache_write":0},"wall_minutes":49.5}},"cells":{"add-i-key":{"scores":"6.833 / 7.9","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-production-review-panels-three-seed-20260723/sol-plan-terra-exec-grok-review/add-i-key.patch","judge_samples":{"fable-5":{"mean":6.833,"stddev":0.236,"scores":[6.5,7.0,7.0],"interval":[6.598,7.069],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":7.9,"stddev":0.294,"scores":[7.5,8.0,8.2],"interval":[7.606,8.194],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"web-install":{"scores":"4.667 / 2.833","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-production-review-panels-three-seed-20260723/sol-plan-terra-exec-grok-review/web-install.patch","judge_samples":{"fable-5":{"mean":4.667,"stddev":0.236,"scores":[4.5,4.5,5.0],"interval":[4.431,4.902],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":2.833,"stddev":0.236,"scores":[2.5,3.0,3.0],"interval":[2.598,3.069],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"install":{"scores":"5.5 / 2.567","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-production-review-panels-three-seed-20260723/sol-plan-terra-exec-grok-review/install.patch","judge_samples":{"fable-5":{"mean":5.5,"stddev":0.408,"scores":[6.0,5.0,5.5],"interval":[5.092,5.908],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":2.567,"stddev":0.094,"scores":[2.5,2.5,2.7],"interval":[2.472,2.661],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"fix-tmux":{"scores":"8.5 / 6.267","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-production-review-panels-three-seed-20260723/sol-plan-terra-exec-grok-review/fix-tmux.patch","judge_samples":{"gpt-5.6-sol":{"mean":6.267,"stddev":0.556,"scores":[5.5,6.8,6.5],"interval":[5.711,6.822],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true},"fable-5":{"mean":8.5,"stddev":0.0,"scores":[8.5,8.5,8.5],"interval":[8.5,8.5],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true}},"recovered_artifact":false},"fix-review":{"scores":"5.667 / 5.0","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-production-review-panels-three-seed-20260723/sol-plan-terra-exec-grok-review/fix-review.patch","judge_samples":{"fable-5":{"mean":5.667,"stddev":0.471,"scores":[6.0,6.0,5.0],"interval":[5.195,6.138],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":5.0,"stddev":0.408,"scores":[4.5,5.5,5.0],"interval":[4.592,5.408],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"daemon":{"scores":"4.0 / 2.833","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-production-review-panels-three-seed-20260723/sol-plan-terra-exec-grok-review/daemon.patch","judge_samples":{"fable-5":{"mean":4.0,"stddev":0.0,"scores":[4.0,4.0,4.0],"interval":[4.0,4.0],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":2.833,"stddev":0.236,"scores":[2.5,3.0,3.0],"interval":[2.598,3.069],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false}}},{"id":"sol-plan@xhigh->exec@high+grok-review","label":"Sol plan → Sol execute → Sol + Grok review","reasoning":"Sol xhigh plan · Sol high execute · Sol xhigh + Grok xhigh review","note":"The production-like flagship: Sol plans and executes, then Sol and Grok review.","campaign_id":"v3-production-review-panels-three-seed-20260723","samples_per_cell":3,"raw_evidence_published":false,"patches_published":true,"completed":6,"total":6,"mean_minutes":151.0,"time_sample":6,"cost_per_task_usd":null,"cost_total_usd":null,"cost_sample":0,"known_cost_per_task_usd":42.94215,"known_cost_total_usd":257.6529,"known_cost_sample":6,"normalized_mtokens_per_task":63.156,"token_sample":6,"token_scope":"Sol stages only · Grok reviewer telemetry unavailable","token_totals":{"input":8299523,"output":1045322,"cache_read":369591296,"cache_write":0},"efficiency_by_task":{"add-i-key":{"cost_usd":null,"known_cost_usd":29.0976,"normalized_mtokens":41.088,"tokens":{"input":1129848,"output":117602,"cache_read":39840512,"cache_write":0},"wall_minutes":124.0},"web-install":{"cost_usd":null,"known_cost_usd":62.9825,"normalized_mtokens":96.241,"tokens":{"input":1778971,"output":232431,"cache_read":94229504,"cache_write":0},"wall_minutes":178.7},"install":{"cost_usd":null,"known_cost_usd":40.7612,"normalized_mtokens":58.575,"tokens":{"input":1252824,"output":197840,"cache_read":57123840,"cache_write":0},"wall_minutes":145.1},"fix-tmux":{"cost_usd":null,"known_cost_usd":25.1855,"normalized_mtokens":38.298,"tokens":{"input":770841,"output":87043,"cache_read":37440000,"cache_write":0},"wall_minutes":94.1},"fix-review":{"cost_usd":null,"known_cost_usd":39.4621,"normalized_mtokens":56.678,"tokens":{"input":1400700,"output":163396,"cache_read":55113472,"cache_write":0},"wall_minutes":177.3},"daemon":{"cost_usd":null,"known_cost_usd":60.164,"normalized_mtokens":88.057,"tokens":{"input":1966339,"output":247010,"cache_read":85843968,"cache_write":0},"wall_minutes":186.5}},"cells":{"add-i-key":{"scores":"5.167 / 4.767","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-production-review-panels-three-seed-20260723/sol-plan-sol-exec-sol-grok-review/add-i-key.patch","judge_samples":{"fable-5":{"mean":5.167,"stddev":0.236,"scores":[5.0,5.0,5.5],"interval":[4.931,5.402],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":4.767,"stddev":0.556,"scores":[5.3,5.0,4.0],"interval":[4.211,5.322],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"web-install":{"scores":"7.0 / 5.833","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-production-review-panels-three-seed-20260723/sol-plan-sol-exec-sol-grok-review/web-install.patch","judge_samples":{"fable-5":{"mean":7.0,"stddev":0.408,"scores":[7.5,6.5,7.0],"interval":[6.592,7.408],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":5.833,"stddev":0.287,"scores":[5.8,5.5,6.2],"interval":[5.547,6.12],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"install":{"scores":"5.667 / 4.733","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-production-review-panels-three-seed-20260723/sol-plan-sol-exec-sol-grok-review/install.patch","judge_samples":{"fable-5":{"mean":5.667,"stddev":0.471,"scores":[6.0,5.0,6.0],"interval":[5.195,6.138],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":4.733,"stddev":0.573,"scores":[4.8,5.4,4.0],"interval":[4.16,5.307],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"fix-tmux":{"scores":"9.0 / 7.833","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-production-review-panels-three-seed-20260723/sol-plan-sol-exec-sol-grok-review/fix-tmux.patch","judge_samples":{"fable-5":{"mean":9.0,"stddev":0.0,"scores":[9.0,9.0,9.0],"interval":[9.0,9.0],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":7.833,"stddev":0.236,"scores":[8.0,7.5,8.0],"interval":[7.598,8.069],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"fix-review":{"scores":"7.833 / 7.167","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-production-review-panels-three-seed-20260723/sol-plan-sol-exec-sol-grok-review/fix-review.patch","judge_samples":{"fable-5":{"mean":7.833,"stddev":0.471,"scores":[7.5,7.5,8.5],"interval":[7.362,8.305],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":7.167,"stddev":0.785,"scores":[6.3,7.0,8.2],"interval":[6.382,7.951],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"daemon":{"scores":"5.667 / 5.267","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-production-review-panels-three-seed-20260723/sol-plan-sol-exec-sol-grok-review/daemon.patch","judge_samples":{"fable-5":{"mean":5.667,"stddev":0.471,"scores":[5.0,6.0,6.0],"interval":[5.195,6.138],"same_family":false,"reasoning_effort":"xhigh","reasoning_effort_explicit":true},"gpt-5.6-sol":{"mean":5.267,"stddev":0.881,"scores":[6.5,4.8,4.5],"interval":[4.386,6.147],"same_family":true,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false}}},{"id":"all-deepseek-v4-pro-0813@xhigh","label":"DeepSeek V4 Pro 0813 xhigh","reasoning":"DeepSeek V4 Pro 0813 xhigh throughout","note":"DeepSeek V4 Pro plans, executes, and reviews through Pi/OpenRouter.","campaign_id":"v3-deepseek-v4-0813-20260813-r2","samples_per_cell":3,"raw_evidence_published":false,"patches_published":true,"completed":6,"total":6,"mean_minutes":120.0,"time_sample":5,"cost_per_task_usd":0.55,"cost_total_usd":2.7732,"cost_sample":5,"normalized_mtokens_per_task":50.754,"token_sample":5,"token_totals":{"input":1893362,"output":1196341,"cache_read":250681600,"cache_write":0},"efficiency_by_task":{"add-i-key":{"cost_usd":0.1272,"normalized_mtokens":3.905,"tokens":{"input":131886,"output":64867,"cache_read":3708672,"cache_write":0},"wall_minutes":36.2},"web-install":{"cost_usd":0.9637,"normalized_mtokens":103.633,"tokens":{"input":610144,"output":374917,"cache_read":102647808,"cache_write":0},"wall_minutes":146.2},"install":{"cost_usd":0.6331,"normalized_mtokens":67.776,"tokens":{"input":368912,"output":263466,"cache_read":67143680,"cache_write":0},"wall_minutes":110.4},"fix-tmux":{"cost_usd":null,"normalized_mtokens":null,"tokens":null,"wall_minutes":null},"fix-review":{"cost_usd":0.4238,"normalized_mtokens":27.539,"tokens":{"input":342854,"output":203184,"cache_read":26993280,"cache_write":0},"wall_minutes":114.5},"daemon":{"cost_usd":0.6254,"normalized_mtokens":50.918,"tokens":{"input":439566,"output":289907,"cache_read":50188160,"cache_write":0},"wall_minutes":192.9}},"cells":{"add-i-key":{"scores":"9.0 / 1.0","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/all-deepseek-v4-pro-0813-xhigh/add-i-key.patch","judge_samples":{"gpt-5.6-sol":{"mean":1.0,"stddev":0.0,"scores":[1.0,1.0,1.0],"interval":[1.0,1.0],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true},"fable-5":{"mean":9.0,"stddev":0.0,"scores":[9.0,9.0,9.0],"interval":[9.0,9.0],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false}},"recovered_artifact":false},"web-install":{"scores":"8.167 / 2.933","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/all-deepseek-v4-pro-0813-xhigh/web-install.patch","judge_samples":{"gpt-5.6-sol":{"mean":2.933,"stddev":0.094,"scores":[2.8,3.0,3.0],"interval":[2.839,3.028],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true},"fable-5":{"mean":8.167,"stddev":0.236,"scores":[8.0,8.0,8.5],"interval":[7.931,8.402],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false}},"recovered_artifact":false},"install":{"scores":"7.167 / 2.167","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/all-deepseek-v4-pro-0813-xhigh/install.patch","judge_samples":{"gpt-5.6-sol":{"mean":2.167,"stddev":0.236,"scores":[2.5,2.0,2.0],"interval":[1.931,2.402],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true},"fable-5":{"mean":7.167,"stddev":0.471,"scores":[7.5,7.5,6.5],"interval":[6.695,7.638],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false}},"recovered_artifact":false},"fix-tmux":{"scores":"5.5 / 5.833","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/all-deepseek-v4-pro-0813-xhigh/fix-tmux.patch","judge_samples":{"fable-5":{"mean":5.5,"stddev":0.816,"scores":[6.5,4.5,5.5],"interval":[4.684,6.316],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":5.833,"stddev":0.471,"scores":[5.5,5.5,6.5],"interval":[5.362,6.305],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"fix-review":{"scores":"8.0 / 3.5","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/all-deepseek-v4-pro-0813-xhigh/fix-review.patch","judge_samples":{"gpt-5.6-sol":{"mean":3.5,"stddev":0.408,"scores":[3.0,3.5,4.0],"interval":[3.092,3.908],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true},"fable-5":{"mean":8.0,"stddev":0.0,"scores":[8.0,8.0,8.0],"interval":[8.0,8.0],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false}},"recovered_artifact":false},"daemon":{"scores":"8.5 / 4.0","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/all-deepseek-v4-pro-0813-xhigh/daemon.patch","judge_samples":{"gpt-5.6-sol":{"mean":4.0,"stddev":0.408,"scores":[4.0,3.5,4.5],"interval":[3.592,4.408],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true},"fable-5":{"mean":8.5,"stddev":0.0,"scores":[8.5,8.5,8.5],"interval":[8.5,8.5],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false}},"recovered_artifact":false}}},{"id":"deepseek-pro@xhigh->flash@xhigh-exec+pro-review","label":"DeepSeek V4 Pro plan → V4 Flash execute → V4 Pro review","reasoning":"DeepSeek V4 Pro 0813 xhigh plan/review · V4 Flash 0731 xhigh execute","note":"DeepSeek V4 Pro plans and reviews while the dated V4 Flash model executes, all through Pi/OpenRouter.","campaign_id":"v3-deepseek-v4-0813-20260813-r2","samples_per_cell":3,"raw_evidence_published":false,"patches_published":true,"completed":6,"total":6,"mean_minutes":118.8,"time_sample":6,"cost_per_task_usd":0.6,"cost_total_usd":3.5895,"cost_sample":6,"normalized_mtokens_per_task":34.774,"token_sample":6,"token_totals":{"input":3593461,"output":1216444,"cache_read":203833344,"cache_write":0},"efficiency_by_task":{"add-i-key":{"cost_usd":0.16,"normalized_mtokens":4.516,"tokens":{"input":153388,"output":99886,"cache_read":4262272,"cache_write":0},"wall_minutes":41.9},"web-install":{"cost_usd":1.1465,"normalized_mtokens":74.559,"tokens":{"input":1522641,"output":301056,"cache_read":72735360,"cache_write":0},"wall_minutes":179.9},"install":{"cost_usd":0.7142,"normalized_mtokens":43.399,"tokens":{"input":559075,"output":232506,"cache_read":42606976,"cache_write":0},"wall_minutes":113.0},"fix-tmux":{"cost_usd":0.2096,"normalized_mtokens":7.448,"tokens":{"input":394426,"output":102405,"cache_read":6951296,"cache_write":0},"wall_minutes":96.6},"fix-review":{"cost_usd":0.633,"normalized_mtokens":38.425,"tokens":{"input":408841,"output":242152,"cache_read":37774208,"cache_write":0},"wall_minutes":167.3},"daemon":{"cost_usd":0.7262,"normalized_mtokens":40.297,"tokens":{"input":555090,"output":238439,"cache_read":39503232,"cache_write":0},"wall_minutes":114.1}},"cells":{"add-i-key":{"scores":"9.0 / 0.833","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/deepseek-v4-pro-plan-flash-execute-pro-review/add-i-key.patch","judge_samples":{"gpt-5.6-sol":{"mean":0.833,"stddev":0.236,"scores":[0.5,1.0,1.0],"interval":[0.598,1.069],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true},"fable-5":{"mean":9.0,"stddev":0.0,"scores":[9.0,9.0,9.0],"interval":[9.0,9.0],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false}},"recovered_artifact":false},"web-install":{"scores":"3.833 / 2.933","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/deepseek-v4-pro-plan-flash-execute-pro-review/web-install.patch","judge_samples":{"fable-5":{"mean":3.833,"stddev":0.236,"scores":[3.5,4.0,4.0],"interval":[3.598,4.069],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":2.933,"stddev":0.094,"scores":[3.0,2.8,3.0],"interval":[2.839,3.028],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"install":{"scores":"5.667 / 2.5","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/deepseek-v4-pro-plan-flash-execute-pro-review/install.patch","judge_samples":{"fable-5":{"mean":5.667,"stddev":0.471,"scores":[6.0,5.0,6.0],"interval":[5.195,6.138],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":2.5,"stddev":0.408,"scores":[2.5,2.0,3.0],"interval":[2.092,2.908],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"fix-tmux":{"scores":"8.5 / 5.5","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/deepseek-v4-pro-plan-flash-execute-pro-review/fix-tmux.patch","judge_samples":{"gpt-5.6-sol":{"mean":5.5,"stddev":0.408,"scores":[6.0,5.0,5.5],"interval":[5.092,5.908],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true},"fable-5":{"mean":8.5,"stddev":0.0,"scores":[8.5,8.5,8.5],"interval":[8.5,8.5],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false}},"recovered_artifact":false},"fix-review":{"scores":"3.5 / 3.0","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/deepseek-v4-pro-plan-flash-execute-pro-review/fix-review.patch","judge_samples":{"fable-5":{"mean":3.5,"stddev":0.408,"scores":[3.5,3.0,4.0],"interval":[3.092,3.908],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":3.0,"stddev":0.0,"scores":[3.0,3.0,3.0],"interval":[3.0,3.0],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"daemon":{"scores":"7.5 / 5.267","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/deepseek-v4-pro-plan-flash-execute-pro-review/daemon.patch","judge_samples":{"fable-5":{"mean":7.5,"stddev":0.408,"scores":[8.0,7.0,7.5],"interval":[7.092,7.908],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":5.267,"stddev":0.377,"scores":[5.8,5.0,5.0],"interval":[4.89,5.644],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false}}},{"id":"all-ox-alpha@high","label":"GLM 5.3 Flash (0x Alpha) via Pi high","reasoning":"GLM 5.3 Flash (0x Alpha) high throughout via Pi/OpenRouter","note":"GLM 5.3 Flash (0x Alpha) owns planning, implementation, and review through Pi at explicit high reasoning; pre-execution plan review is disabled.","campaign_id":"v3-pi-ox-alpha-high-20260825-r3","samples_per_cell":3,"raw_evidence_published":false,"patches_published":true,"completed":6,"total":6,"mean_minutes":105.7,"time_sample":4,"cost_per_task_usd":null,"cost_total_usd":null,"cost_sample":0,"normalized_mtokens_per_task":19.838,"token_sample":6,"token_totals":{"input":5910682,"output":536749,"cache_read":112580672,"cache_write":0},"efficiency_by_task":{"install":{"cost_usd":null,"normalized_mtokens":29.776,"tokens":{"input":1157836,"output":134357,"cache_read":28483392,"cache_write":0},"wall_minutes":160.6},"fix-tmux":{"cost_usd":null,"normalized_mtokens":3.348,"tokens":{"input":467467,"output":32366,"cache_read":2847808,"cache_write":0},"wall_minutes":99.1},"daemon":{"cost_usd":null,"normalized_mtokens":17.794,"tokens":{"input":1372838,"output":69838,"cache_read":16351616,"cache_write":0},"wall_minutes":null},"add-i-key":{"cost_usd":null,"normalized_mtokens":5.392,"tokens":{"input":651563,"output":64589,"cache_read":4676032,"cache_write":0},"wall_minutes":51.2},"web-install":{"cost_usd":null,"normalized_mtokens":52.999,"tokens":{"input":1342766,"output":149583,"cache_read":51506880,"cache_write":0},"wall_minutes":null},"fix-review":{"cost_usd":null,"normalized_mtokens":9.719,"tokens":{"input":918212,"output":86016,"cache_read":8714944,"cache_write":0},"wall_minutes":111.8}},"cells":{"install":{"scores":"6.833 / 2.167","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-pi-ox-alpha-high-20260825-r3/all-ox-alpha-high/install.patch","judge_samples":{"fable-5":{"mean":6.833,"stddev":0.236,"scores":[7.0,7.0,6.5],"interval":[6.598,7.069],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":2.167,"stddev":0.236,"scores":[2.5,2.0,2.0],"interval":[1.931,2.402],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"fix-tmux":{"scores":"8.167 / 5.5","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-pi-ox-alpha-high-20260825-r3/all-ox-alpha-high/fix-tmux.patch","judge_samples":{"fable-5":{"mean":8.167,"stddev":0.236,"scores":[8.0,8.5,8.0],"interval":[7.931,8.402],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":5.5,"stddev":0.408,"scores":[6.0,5.5,5.0],"interval":[5.092,5.908],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"daemon":{"scores":"6.833 / 3.0","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-pi-ox-alpha-high-20260825-r3/all-ox-alpha-high/daemon.patch","judge_samples":{"fable-5":{"mean":6.833,"stddev":0.236,"scores":[6.5,7.0,7.0],"interval":[6.598,7.069],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":3.0,"stddev":0.0,"scores":[3.0,3.0,3.0],"interval":[3.0,3.0],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":true},"add-i-key":{"scores":"6.833 / 6.0","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-pi-ox-alpha-high-20260825-r3/all-ox-alpha-high/add-i-key.patch","judge_samples":{"fable-5":{"mean":6.833,"stddev":0.471,"scores":[7.5,6.5,6.5],"interval":[6.362,7.305],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":6.0,"stddev":0.408,"scores":[5.5,6.5,6.0],"interval":[5.592,6.408],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"web-install":{"scores":"6.0 / 3.767","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-pi-ox-alpha-high-20260825-r3/all-ox-alpha-high/web-install.patch","judge_samples":{"fable-5":{"mean":6.0,"stddev":0.0,"scores":[6.0,6.0,6.0],"interval":[6.0,6.0],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":3.767,"stddev":0.205,"scores":[4.0,3.8,3.5],"interval":[3.561,3.972],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":true},"fix-review":{"scores":"3.167 / 2.333","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-pi-ox-alpha-high-20260825-r3/all-ox-alpha-high/fix-review.patch","judge_samples":{"fable-5":{"mean":3.167,"stddev":0.236,"scores":[3.5,3.0,3.0],"interval":[2.931,3.402],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":2.333,"stddev":0.236,"scores":[2.0,2.5,2.5],"interval":[2.098,2.569],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false}}},{"id":"all-ox-alpha@max","label":"GLM 5.3 Flash (0x Alpha) via Pi max","reasoning":"GLM 5.3 Flash (0x Alpha) max throughout via Pi/OpenRouter","note":"GLM 5.3 Flash (0x Alpha) owns planning, implementation, and review through Pi at explicit max reasoning; pre-execution plan review is disabled.","campaign_id":"v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed","samples_per_cell":3,"raw_evidence_published":false,"patches_published":true,"completed":6,"total":6,"mean_minutes":95.5,"time_sample":5,"cost_per_task_usd":6.9381,"cost_total_usd":41.6283,"cost_sample":6,"normalized_mtokens_per_task":33.343,"token_sample":6,"token_totals":{"input":4481726,"output":768386,"cache_read":194807808,"cache_write":0},"efficiency_by_task":{"add-i-key":{"cost_usd":3.3947,"normalized_mtokens":16.316,"tokens":{"input":263419,"output":90393,"cache_read":15962368,"cache_write":0},"wall_minutes":44.6},"web-install":{"cost_usd":7.6229,"normalized_mtokens":38.153,"tokens":{"input":448416,"output":145401,"cache_read":37559232,"cache_write":0},"wall_minutes":4.5},"install":{"cost_usd":15.4779,"normalized_mtokens":71.713,"tokens":{"input":2690990,"output":176425,"cache_read":68845440,"cache_write":0},"wall_minutes":233.8},"fix-tmux":{"cost_usd":1.9862,"normalized_mtokens":8.619,"tokens":{"input":228883,"output":91708,"cache_read":8298048,"cache_write":0},"wall_minutes":65.3},"fix-review":{"cost_usd":6.6754,"normalized_mtokens":33.903,"tokens":{"input":321255,"output":115439,"cache_read":33466112,"cache_write":0},"wall_minutes":null},"daemon":{"cost_usd":6.4712,"normalized_mtokens":31.354,"tokens":{"input":528763,"output":149020,"cache_read":30676608,"cache_write":0},"wall_minutes":129.2}},"cells":{"add-i-key":{"scores":"8.667 / 4.667","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed/all-ox-alpha-max/add-i-key.patch","judge_samples":{"gpt-5.6-sol":{"mean":4.667,"stddev":0.236,"scores":[5.0,4.5,4.5],"interval":[4.431,4.902],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true},"fable-5":{"mean":8.667,"stddev":0.471,"scores":[9.0,8.0,9.0],"interval":[8.195,9.138],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false}},"recovered_artifact":false},"web-install":{"scores":"4.333 / 3.1","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed/all-ox-alpha-max/web-install.patch","judge_samples":{"fable-5":{"mean":4.333,"stddev":0.471,"scores":[4.0,5.0,4.0],"interval":[3.862,4.805],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":3.1,"stddev":0.294,"scores":[2.8,3.5,3.0],"interval":[2.806,3.394],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false},"install":{"scores":"6.0 / 1.667","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed/all-ox-alpha-max/install.patch","judge_samples":{"gpt-5.6-sol":{"mean":1.667,"stddev":0.236,"scores":[1.5,2.0,1.5],"interval":[1.431,1.902],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true},"fable-5":{"mean":6.0,"stddev":0.0,"scores":[6.0,6.0,6.0],"interval":[6.0,6.0],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false}},"recovered_artifact":false},"fix-tmux":{"scores":"8.667 / 5.167","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed/all-ox-alpha-max/fix-tmux.patch","judge_samples":{"gpt-5.6-sol":{"mean":5.167,"stddev":0.236,"scores":[5.5,5.0,5.0],"interval":[4.931,5.402],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true},"fable-5":{"mean":8.667,"stddev":0.236,"scores":[8.5,9.0,8.5],"interval":[8.431,8.902],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false}},"recovered_artifact":false},"fix-review":{"scores":"4.0 / 2.5","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed/all-ox-alpha-max/fix-review.patch","judge_samples":{"fable-5":{"mean":4.0,"stddev":0.0,"scores":[4.0,4.0,4.0],"interval":[4.0,4.0],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":2.5,"stddev":0.408,"scores":[3.0,2.5,2.0],"interval":[2.092,2.908],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":true},"daemon":{"scores":"7.333 / 4.333","result_url":"/bench/results.json","patch_url":"/bench/patches/v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed/all-ox-alpha-max/daemon.patch","judge_samples":{"fable-5":{"mean":7.333,"stddev":0.236,"scores":[7.5,7.0,7.5],"interval":[7.098,7.569],"same_family":false,"reasoning_effort":"unspecified","reasoning_effort_explicit":false},"gpt-5.6-sol":{"mean":4.333,"stddev":0.236,"scores":[4.5,4.5,4.0],"interval":[4.098,4.569],"same_family":false,"reasoning_effort":"ultra","reasoning_effort_explicit":true}},"recovered_artifact":false}}}],"evidence":{"readme":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/README.md","manifest":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/manifest.json","results":"https://github.com/ivankuznetsov/hive-bench/blob/main/runs/v2-ce/results.json","cells":"https://github.com/ivankuznetsov/hive-bench/tree/main/runs/v2-ce/cells","followup_results":"/bench/results.json","followup_raw_evidence_published":false,"followup_patches":"/bench/patches/v3-mixed-workflows-three-seed-20260713/","followup_patches_published":true,"production_review_panel_patches":"/bench/patches/v3-production-review-panels-three-seed-20260723/","production_review_panel_patches_published":true,"deepseek_patches":"/bench/patches/v3-deepseek-v4-0813-20260813-r2/","deepseek_patches_published":true,"ox_alpha_pi_patches":"/bench/patches/v3-pi-ox-alpha-high-20260825-r3/","ox_alpha_pi_patches_published":true,"ox_alpha_opencode_patches":"/bench/patches/v3-opencode-ox-alpha-high-20260825-r3/","ox_alpha_opencode_patches_published":true,"ox_alpha_pi_max_patches":"/bench/patches/v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed/","ox_alpha_pi_max_patches_published":true},"notes":["All 90 generation cells completed. Every objective gate is no_gate in all six active campaigns, so scores are judge evidence, not test-pass rates.","The original 36-cell campaign uses one sample per judge; the 54 cells across the five later active three-sample campaigns use three samples per judge. All six active campaigns ran adversarial deliberation across all 90 cells.","Discussion final is a separate one-shot diagnostic, not an adjustment to the independent leaderboard scores. It covers 180/180 judge decisions. The original campaign reused its exact published verdicts and locally recovered rationales for round two; the five later active campaigns freshly re-graded round one.","GLM 5.3 Flash (0x Alpha) via Pi high preserves tokens for 6/6 cells and wall time for 4/6. Cost remains unknown rather than inferred.","GLM 5.3 Flash (0x Alpha) via Pi max preserves price and token telemetry for 6/6 cells and wall time for 5/6; the recovered fix-review artifact has no wall-time sample.","The historical OpenCode GLM 5.3 Flash (0x Alpha) campaign remains withdrawn because its profile lacked Bash capability parity and its target exposed held-out history; its invalidated patch artifacts remain available for audit.","Efficiency uses each campaign's serialized generation telemetry. DeepSeek's Pi/OpenRouter path preserved complete price and token telemetry for 11/12 cells and wall time for 11/12; missing values remain unknown.","Five Sol → Terra → Sol cells have complete recomputed API-equivalent costs; Grok workflows publish known-provider subtotals because Grok usage telemetry is unavailable.","The mixed Opus → Codex estimate is $115.5829 across six cells: $67.3351 Codex 5.5, $47.6319 Opus 4.8, and $0.6159 Haiku utility calls, or $19.26 per task on average.","Grok workflows publish known-provider token and API-equivalent cost subtotals. They exclude Grok, remain partial, and never impute its missing telemetry.","Wall-time means use only recorded samples. The production-review-panel campaign retains wall time for all 12 cells.","Costs use the benchmark's versioned usual-tier tables and are API-equivalent estimates, not subscription invoices; judge usage is excluded.","All 54 active three-sample campaign patches are public on this site; raw provider streams, logs, target clones, and auth material remain unpublished. Withdrawn OpenCode GLM 5.3 Flash (0x Alpha) patches remain available only as explicitly invalidated audit artifacts."],"campaigns":[{"id":"v2-ce","corpus_version":"v2-ce","cells":36,"candidates":6,"judge_samples_per_cell":1,"sol_reasoning_effort":"xhigh","deliberation_complete":true,"raw_evidence_published":true},{"id":"v3-mixed-workflows-three-seed-20260713","corpus_version":"v3-mixed-workflows-followup-20260713","cells":18,"candidates":3,"judge_samples_per_cell":3,"sol_reasoning_effort":"ultra","deliberation_complete":true,"raw_evidence_published":false,"patches_published":true},{"id":"v3-production-review-panels-three-seed-20260723","corpus_version":"v3-production-review-panels-20260723","cells":12,"candidates":2,"judge_samples_per_cell":3,"sol_reasoning_effort":"ultra","deliberation_complete":true,"raw_evidence_published":false,"patches_published":true},{"id":"v3-deepseek-v4-0813-20260813-r2","corpus_version":"v3-deepseek-v4-0813-20260813-r2","cells":12,"candidates":2,"judge_samples_per_cell":3,"fable_reasoning_effort":"unspecified","sol_reasoning_effort":"ultra","deliberation_complete":true,"raw_evidence_published":false,"patches_published":true},{"id":"v3-pi-ox-alpha-high-20260825-r3","corpus_version":"v3-pi-ox-alpha-high-20260825-r3","cells":6,"candidates":1,"judge_samples_per_cell":3,"fable_reasoning_effort":"unspecified","sol_reasoning_effort":"ultra","deliberation_complete":true,"raw_evidence_published":false,"patches_published":true},{"id":"v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed","corpus_version":"v3-pi-ox-alpha-glm-5-3-flash-max-20260826-r6-sealed","cells":6,"candidates":1,"judge_samples_per_cell":3,"fable_reasoning_effort":"unspecified","sol_reasoning_effort":"ultra","deliberation_complete":true,"raw_evidence_published":false,"patches_published":true}],"withdrawn_campaigns":[{"id":"v3-opencode-ox-alpha-high-20260825-r3","candidate_id":"all-ox-alpha-opencode@high","label":"GLM 5.3 Flash (0x Alpha) via OpenCode high","withdrawn_at":"2026-08-26T16:20:00Z","disposition":"excluded from ranking and active coverage","reason":"The historical target clone exposed full repository history, including the exact held-out merge commit, and allowed ordinary GitHub egress. The resolved OpenCode profile allowed only Read, Write, and Edit, without Bash. All six stages reached Hive execute-complete, but the model could not inspect through a shell or run tests like Pi.","copying_evidence":"No positive gold-copying evidence was found, but retained OpenCode events redact tool payloads, so a negative lookup audit is not possible.","artifact_path":"/bench/patches/v3-opencode-ox-alpha-high-20260825-r3/"}]}
