[{"id":"caching-turns","title":"Cache counters per turn (mean over sessions)","columns":{"$k":["key","label","unit"],"$r":[["config","Configuration","text"],["turn","Turn","count"],["sessions","Sessions","count"],["correct","Exact answers","text"],["input","Input tokens (all)","tokens"],["read","Read from cache","tokens"],["written","Written to cache","text"],["share","Read share","rate"],["medianTotal","Median time (s)","seconds"],["withCache","USD with cache (calculation)","usd"],["without","USD without cache (calculation)","usd"]]},"rows":{"$k":["config","turn","sessions","correct","input","read","written","share","medianTotal","withCache","without"],"$r":[["Claude Sonnet 5.5 · Claude Code",1,3,"3/3",7831,1463,"6,366",0.1868,1.64,0.025792,0.015693],["Claude Sonnet 5.5 · Claude Code",2,3,"3/3",7889,7829,"58",0.9924,1.54,0.001862,0.015839],["Claude Sonnet 5.5 · Claude Code",3,3,"3/3",7970,7887,"81",0.9896,1.53,0.001955,0.015991],["Claude Sonnet 5.5 · Claude Code",4,3,"3/3",8032,7968,"62",0.992,5.2,0.009576,0.023795],["Claude Sonnet 5.5 · Claude Code",5,3,"3/3",8861,8030,"829",0.9074,1.77,0.005816,0.018613],["Claude Opus 5.5 · Claude Code",1,3,"3/3",7828,1463,"6,363",0.1869,1.9,0.051265,0.031372],["Claude Opus 5.5 · Claude Code",2,3,"3/3",7886,7826,"58",0.9924,1.63,0.002637,0.032144],["Claude Opus 5.5 · Claude Code",3,3,"3/3",7991,7884,"105",0.9866,1.81,0.002965,0.032504],["Claude Opus 5.5 · Claude Code",4,3,"3/3",8075,7989,"84",0.9893,6.09,0.018471,0.048493],["Claude Opus 5.5 · Claude Code",5,3,"3/3",8941,8073,"866",0.9031,2.09,0.009681,0.036896],["GPT-6.1 Sol (medium) · Codex CLI",1,3,"3/3",18774,10411,"not reported",0.5545,3.56,null,null],["GPT-6.1 Sol (medium) · Codex CLI",2,3,"3/3",18803,18560,"not reported",0.9871,2.39,null,null],["GPT-6.1 Sol (medium) · Codex CLI",3,3,"3/3",18848,18645,"not reported",0.9892,3.11,null,null],["GPT-6.1 Sol (medium) · Codex CLI",4,3,"3/3",18880,18688,"not reported",0.9898,5.97,null,null],["GPT-6.1 Sol (medium) · Codex CLI",5,3,"3/3",19097,18688,"not reported",0.9786,2.38,null,null]]}},{"id":"consistency-cells","title":"Every consistency cell (10 repetitions each)","columns":{"$k":["key","label","unit"],"$r":[["config","Configuration","text"],["prompt","Prompt","text"],["strict","Strict passes","text"],["ci","95% interval","text"],["formatMisses","Format misses","count"],["wrong","Wrong answers","count"],["distinct","Distinct answers","count"],["distinctRaw","Distinct raw replies","count"],["medianTotal","Median time (s)","seconds"],["iqr","Interquartile range (s)","seconds"],["cv","Coefficient of variation","score"]]},"rows":{"$k":["config","prompt","strict","ci","formatMisses","wrong","distinct","distinctRaw","medianTotal","iqr","cv"],"$r":[["Claude Haiku 4.5 · Claude Code","Exact number","0/10","0% to 28%",0,10,1,10,5.06,1.12,0.13],["Claude Haiku 4.5 · Claude Code","JSON object","1/10","2% to 40%",9,0,1,3,7.03,2.21,0.28],["Claude Haiku 4.5 · Claude Code","Code fix","10/10","72% to 100%",0,0,6,6,5.95,1.09,0.13],["Claude Sonnet 5.5 · Claude Code","Exact number","10/10","72% to 100%",0,0,1,1,6.89,0.95,0.09],["Claude Sonnet 5.5 · Claude Code","JSON object","10/10","72% to 100%",0,0,1,1,2.89,0.54,0.28],["Claude Sonnet 5.5 · Claude Code","Code fix","10/10","72% to 100%",0,0,3,3,2.67,1.17,0.26],["GPT-6.1 Sol (medium) · Codex CLI","Exact number","10/10","72% to 100%",0,0,1,1,13.38,1.53,0.14],["GPT-6.1 Sol (medium) · Codex CLI","JSON object","10/10","72% to 100%",0,0,1,1,6.42,1.8,0.17],["GPT-6.1 Sol (medium) · Codex CLI","Code fix","10/10","72% to 100%",0,0,6,6,11.29,2.25,0.15]]}}]