!function(){try{var e="undefined"!=typeof window?window:"undefined"!=typeof global?global:"undefined"!=typeof globalThis?globalThis:"undefined"!=typeof self?self:{},n=(new e.Error).stack;n&&(e._posthogChunkIds=e._posthogChunkIds||{},e._posthogChunkIds[n]="01a08fa9-635e-7b13-9b2d-0367079a54a2")}catch(e){}}();(globalThis.TURBOPACK||(globalThis.TURBOPACK=[])).push(["object"==typeof document?document.currentScript:void 0,290124,e=>{"use strict";var t=e.i(222666),a=e.i(505962),r=e.i(553215),s=e.i(252358),l=e.i(378039);function o(e){return{question:e.question,answer:e.answer}}e.s(["BenchmarkFaqList",0,function(e){let n,i,c,d=(0,a.c)(8),{benchmark:u,faqs:m}=e,h=(0,s.usePosthogClient)();return d[0]!==m?(n=m.map(o),d[0]=m,d[1]=n):n=d[1],d[2]!==u||d[3]!==h?(i=e=>h.capture(r.PostHogEvent.ExpandBenchmarkFaq,{benchmark:u,question:e.question}),d[2]=u,d[3]=h,d[4]=i):i=d[4],d[5]!==n||d[6]!==i?(c=(0,t.jsx)(l.FAQAccordion,{items:n,onItemExpand:i}),d[5]=n,d[6]=i,d[7]=c):c=d[7],c}])},460867,e=>{"use strict";var t=e.i(222666),a=e.i(505962),r=e.i(452473);e.s(["BenchmarkChartTooltip",0,function(e){let s,l,o,n,i=(0,a.c)(10),{bodyClassName:c,children:d,label:u}=e;return i[0]!==u?(s=(0,t.jsx)("div",{className:"font-medium text-foreground",children:u}),i[0]=u,i[1]=s):s=i[1],i[2]!==c?(l=(0,r.cn)("mt-1 text-muted-foreground",c),i[2]=c,i[3]=l):l=i[3],i[4]!==d||i[5]!==l?(o=(0,t.jsx)("div",{className:l,children:d}),i[4]=d,i[5]=l,i[6]=o):o=i[6],i[7]!==s||i[8]!==o?(n=(0,t.jsxs)("div",{className:"rounded-md border border-border bg-popover px-3 py-2 text-xs shadow-md",children:[s,o]}),i[7]=s,i[8]=o,i[9]=n):n=i[9],n}])},627413,e=>{"use strict";var t=e.i(222666),a=e.i(505962),r=e.i(452473),s=e.i(970654),l=e.i(956506),o=e.i(759024),n=e.i(133974),i=e.i(379684),c=e.i(648156),d=e.i(779100),u=e.i(193975),m=e.i(639222),h=e.i(270401),x=e.i(265716),g=e.i(714170);let p=["accuracy","cost","duration","tokens"];function f(e){let r,s,l,o,i,c,m,h,x,p,f,v,b,j=(0,a.c)(35),{provider:k}=e,y=(0,d.useLocale)();return j[0]===Symbol.for("react.memo_cache_sentinel")?(r=(0,t.jsx)(n.TableCell,{}),j[0]=r):r=j[0],j[1]!==k.providerName?(s=(0,t.jsx)(n.TableRowHead,{children:(0,t.jsx)("span",{className:"pl-9 text-muted-foreground",children:k.providerName})}),j[1]=k.providerName,j[2]=s):s=j[2],j[3]!==y||j[4]!==k.accuracy?(l=(0,g.formatPercent)(y,k.accuracy),j[3]=y,j[4]=k.accuracy,j[5]=l):l=j[5],j[6]!==l?(o=(0,t.jsx)(n.TableCell,{className:"text-right",children:(0,t.jsx)("span",{className:"w-12 tabular-nums text-muted-foreground",children:l})}),j[6]=l,j[7]=o):o=j[7],j[8]!==y||j[9]!==k.accuracyStdDev?(i=null===k.accuracyStdDev?"--":`\xb1${(0,u.formatNativeFixed)(100*k.accuracyStdDev,{locale:y,fractionDigits:1})??"—"}pp`,j[8]=y,j[9]=k.accuracyStdDev,j[10]=i):i=j[10],j[11]!==i?(c=(0,t.jsx)(n.TableCell,{className:"text-right tabular-nums text-muted-foreground",children:i}),j[11]=i,j[12]=c):c=j[12],j[13]!==y||j[14]!==k.avgCostPerTask?(m=(0,g.formatCost)(y,k.avgCostPerTask),j[13]=y,j[14]=k.avgCostPerTask,j[15]=m):m=j[15],j[16]!==m?(h=(0,t.jsx)(n.TableCell,{className:"text-right tabular-nums text-muted-foreground",children:m}),j[16]=m,j[17]=h):h=j[17],j[18]!==y||j[19]!==k.avgDurationPerTaskMs?(x=(0,g.formatDurationMs)(y,k.avgDurationPerTaskMs),j[18]=y,j[19]=k.avgDurationPerTaskMs,j[20]=x):x=j[20],j[21]!==x?(p=(0,t.jsx)(n.TableCell,{className:"text-right tabular-nums text-muted-foreground",children:x}),j[21]=x,j[22]=p):p=j[22],j[23]!==y||j[24]!==k.avgOutputTokensPerTask?(f=(0,g.formatTokens)(y,k.avgOutputTokensPerTask),j[23]=y,j[24]=k.avgOutputTokensPerTask,j[25]=f):f=j[25],j[26]!==f?(v=(0,t.jsx)(n.TableCell,{className:"text-right tabular-nums text-muted-foreground",children:f}),j[26]=f,j[27]=v):v=j[27],j[28]!==p||j[29]!==v||j[30]!==s||j[31]!==o||j[32]!==c||j[33]!==h?(b=(0,t.jsxs)(n.TableRow,{className:"bg-muted/40",children:[r,s,o,c,h,p,v]}),j[28]=p,j[29]=v,j[30]=s,j[31]=o,j[32]=c,j[33]=h,j[34]=b):b=j[34],b}function v(e){return e.providers.length>1}function b(e,t){switch(t){case"accuracy":return e.median.accuracy;case"cost":return e.median.avgCostPerTask;case"duration":return e.median.avgDurationPerTaskMs;case"tokens":return e.median.avgOutputTokensPerTask}}e.s(["ProviderRow",0,f,"Tau2AirlineLeaderboard",0,function({caption:e,metricLabel:a="task",models:j}){let k=(0,d.useLocale)(),[y,T]=(0,h.useState)("accuracy"),[P,C]=(0,h.useState)("descending"),[S,N]=(0,h.useState)(new Set),w=(0,h.useMemo)(()=>new Map(j.toSorted((e,t)=>t.median.accuracy-e.median.accuracy).map((e,t)=>[e.modelPermaslug,t+1])),[j]),D=(0,h.useMemo)(()=>(function({models:e,key:t,direction:a}){return e.toSorted((e,r)=>(0,x.compareNullableNumbers)(b(e,t),b(r,t),a))})({models:j,key:y,direction:P}),[j,y,P]);return(0,t.jsx)(l.Card,{className:"overflow-hidden",children:(0,t.jsxs)(n.Table,{compact:!0,sort:{key:y,direction:"ascending"===P?"asc":"desc"},onSortChange:function(e){let t=p.find(t=>t===e?.key);if(null===e||void 0===t){T("accuracy"),C("descending");return}T(t),C("asc"===e.direction?"ascending":"descending")},children:[(0,t.jsx)(n.TableCaption,{className:"sr-only",children:e}),(0,t.jsx)(n.TableHeader,{children:(0,t.jsxs)(n.TableRow,{children:[(0,t.jsx)(n.TableHead,{className:"w-10 text-right",scope:"col",children:"#"}),(0,t.jsx)(n.TableHead,{scope:"col",children:"Model"}),(0,t.jsx)(n.TableHead,{className:"text-right",sortKey:"accuracy",sortFirstDirection:"descending",scope:"col",children:"Accuracy"}),(0,t.jsx)(n.TableHead,{className:"text-right",scope:"col",children:(0,t.jsxs)("span",{className:"inline-flex items-center gap-1",children:["Std dev",(0,t.jsx)(o.InfoTooltip,{children:"Top-level rows show standard deviation across runs using default routing (not pinned to a provider); provider rows show standard deviation across runs pinned to that provider."})]})}),(0,t.jsxs)(n.TableHead,{className:"text-right",sortKey:"cost",scope:"col",children:["Cost / ",a]}),(0,t.jsxs)(n.TableHead,{className:"text-right",sortKey:"duration",scope:"col",children:["Time / ",a]}),(0,t.jsxs)(n.TableHead,{className:"text-right",sortKey:"tokens",scope:"col",children:["Output tok / ",a]})]})}),(0,t.jsx)(n.TableBody,{children:D.map(e=>(0,t.jsxs)(h.Fragment,{children:[(0,t.jsxs)(n.TableRow,{className:v(e)?"cursor-pointer":void 0,onClick:v(e)?()=>{var t;return t=e.modelPermaslug,void N(e=>{let a=new Set(e);return a.has(t)?a.delete(t):a.add(t),a})}:void 0,children:[(0,t.jsx)(n.TableCell,{className:"text-right tabular-nums text-muted-foreground",children:w.get(e.modelPermaslug)}),(0,t.jsx)(n.TableRowHead,{className:"w-80 max-w-80 overflow-hidden",children:(0,t.jsx)("div",{className:"min-w-0 max-w-80",children:(0,t.jsxs)("div",{className:"flex min-w-0 items-center gap-2",children:[v(e)?(0,t.jsx)(m.ChevronRightIcon,{className:(0,r.cn)("size-3.5 shrink-0 text-muted-foreground transition-transform",S.has(e.modelPermaslug)&&"rotate-90")}):(0,t.jsx)("span",{"aria-hidden":"true",className:"size-3.5 shrink-0"}),e.modelSlug?(0,t.jsx)(i.TextLink,{className:"min-w-0 truncate",href:`/${e.modelSlug}`,onClick:e=>e.stopPropagation(),children:e.modelName}):(0,t.jsx)("span",{className:"min-w-0 truncate",children:e.modelName}),e.isParetoOptimal&&(0,t.jsx)(c.Tooltip,{content:"Pareto-optimal: no other model is both cheaper and more accurate. See the Accuracy vs. cost chart.",children:(0,t.jsx)(s.Badge,{className:"shrink-0 cursor-help",variant:"secondary",children:"Pareto"})})]})})}),(0,t.jsx)(n.TableCell,{className:"text-right tabular-nums",children:(0,g.formatPercent)(k,e.median.accuracy)}),(0,t.jsx)(n.TableCell,{className:"text-right tabular-nums text-muted-foreground",children:null===e.accuracyStdDev?"--":`\xb1${(0,u.formatNativeFixed)(100*e.accuracyStdDev,{locale:k,fractionDigits:1})??"—"}pp`}),(0,t.jsx)(n.TableCell,{className:"text-right tabular-nums text-muted-foreground",children:(0,g.formatCost)(k,e.median.avgCostPerTask)}),(0,t.jsx)(n.TableCell,{className:"text-right tabular-nums text-muted-foreground",children:(0,g.formatDurationMs)(k,e.median.avgDurationPerTaskMs)}),(0,t.jsx)(n.TableCell,{className:"text-right tabular-nums text-muted-foreground",children:(0,g.formatTokens)(k,e.median.avgOutputTokensPerTask)})]}),v(e)&&S.has(e.modelPermaslug)&&e.providers.map(e=>(0,t.jsx)(f,{provider:e},e.id))]},e.modelPermaslug))})]})})}])},414374,e=>{"use strict";var t=e.i(222666),a=e.i(505962),r=e.i(536652),s=e.i(657691),l=e.i(553215),o=e.i(528164),n=e.i(956506),i=e.i(389097),c=e.i(557225),d=e.i(145152),u=e.i(379684),m=e.i(20629),h=e.i(252358),x=e.i(779100),g=e.i(270401),p=e.i(307425),f=e.i(5353),v=e.i(4696),b=e.i(249743),j=e.i(537457),k=e.i(487154),y=e.i(429088),T=e.i(287172),P=e.i(612960),C=e.i(460867),S=e.i(265716),N=e.i(714170);function w(e){return null!==e.median.avgDurationPerTaskMs}function D(e,t){return(0,S.compareNullableNumbers)(e.median.avgDurationPerTaskMs,t.median.avgDurationPerTaskMs,"ascending")}function M(e,t){return t.median.accuracy-e.median.accuracy}function A(e){return e.author}function _(e){return{author:e.author,costPerTask:e.median.avgCostPerTask,id:e.modelPermaslug,label:e.modelName,modelSlug:e.modelSlug,accuracyPercent:100*e.median.accuracy,secondsPerTask:null===e.median.avgDurationPerTaskMs?null:e.median.avgDurationPerTaskMs/1e3}}function B(e){return{description:e.author,label:e.modelName,value:e.modelPermaslug}}function K(e){return e.modelPermaslug}function E(e){return e.modelPermaslug}function F(e){let r,s,l,o=(0,a.c)(9),{label:i,model:d,onSelect:m,value:h}=e;return o[0]!==i?(r=(0,t.jsx)("p",{className:"text-2xs font-medium uppercase tracking-widest text-muted-foreground",children:i}),o[0]=i,o[1]=r):r=o[1],o[2]!==d||o[3]!==m||o[4]!==h?(s=null===d?(0,t.jsx)("span",{className:"text-base font-semibold",children:"--"}):(0,t.jsxs)("div",{className:"flex min-w-0 items-center gap-2",children:[(0,t.jsx)(c.AuthorIcon,{author:d.author,className:"size-6",frame:!0}),(0,t.jsxs)("div",{className:"min-w-0",children:[null===d.modelSlug?(0,t.jsx)("p",{className:"truncate text-base font-semibold",children:d.modelName}):(0,t.jsx)(u.TextLink,{className:"block truncate text-base font-semibold",href:`/${d.modelSlug}`,onClick:()=>{null!==d.modelSlug&&m(d.modelSlug)},children:d.modelName}),(0,t.jsx)("p",{className:"text-sm text-muted-foreground",children:h??"--"})]})]}),o[2]=d,o[3]=m,o[4]=h,o[5]=s):s=o[5],o[6]!==r||o[7]!==s?(l=(0,t.jsxs)(n.Card,{className:"flex flex-col gap-2 p-4",children:[r,s]}),o[6]=r,o[7]=s,o[8]=l):l=o[8],l}function R(e){let s,c,d,u,m,x,p,f,v,b,j,k,y,T,C,S=(0,a.c)(50),{benchmark:N,colorByAuthor:w,data:D,dataKey:M,description:A,title:_,valueFormatter:B}=e,K=(0,h.usePosthogClient)(),E=(0,P.useLocalizedRouter)(),[F,R]=(0,g.useState)(!1);S[0]!==N||S[1]!==M||S[2]!==K||S[3]!==E?(s=e=>{null!==e.modelSlug&&(K.capture(l.PostHogEvent.ViewBenchmarkModel,{benchmark:N,chart:M,model_slug:e.modelSlug}),E.push(`/${e.modelSlug}`))},S[0]=N,S[1]=M,S[2]=K,S[3]=E,S[4]=s):s=S[4];let L=s;S[5]!==N||S[6]!==M||S[7]!==K?(c=()=>{R(!0),K.capture(l.PostHogEvent.ExpandBenchmarkComparisonChart,{benchmark:N,chart:M})},S[5]=N,S[6]=M,S[7]=K,S[8]=c):c=S[8];let O=c;S[9]!==_?(d=(0,t.jsx)(n.CardTitle,{children:_}),S[9]=_,S[10]=d):d=S[10];let $=`Expand ${_} chart`;return S[11]===Symbol.for("react.memo_cache_sentinel")?(u=(0,t.jsx)(r.ArrowsPointingOutIcon,{className:"size-4"}),S[11]=u):u=S[11],S[12]!==O||S[13]!==$?(m=(0,t.jsx)(o.Button,{"aria-label":$,className:"-mr-1 -mt-1 shrink-0",onClick:O,size:"icon-sm",variant:"ghost",children:u}),S[12]=O,S[13]=$,S[14]=m):m=S[14],S[15]!==d||S[16]!==m?(x=(0,t.jsxs)("div",{className:"flex items-start justify-between gap-2",children:[d,m]}),S[15]=d,S[16]=m,S[17]=x):x=S[17],S[18]!==A?(p=(0,t.jsx)(n.CardDescription,{className:"mt-1",children:A}),S[18]=A,S[19]=p):p=S[19],S[20]!==w||S[21]!==D||S[22]!==M||S[23]!==L||S[24]!==B?(f=(0,t.jsx)("div",{className:"mt-4",children:(0,t.jsx)(H,{colorByAuthor:w,data:D,dataKey:M,onModelClick:L,valueFormatter:B})}),S[20]=w,S[21]=D,S[22]=M,S[23]=L,S[24]=B,S[25]=f):f=S[25],S[26]!==_?(v=(0,t.jsx)(i.DialogTitle,{children:_}),S[26]=_,S[27]=v):v=S[27],S[28]!==A?(b=(0,t.jsx)(i.DialogDescription,{children:A}),S[28]=A,S[29]=b):b=S[29],S[30]!==v||S[31]!==b?(j=(0,t.jsxs)(i.DialogHeader,{children:[v,b]}),S[30]=v,S[31]=b,S[32]=j):j=S[32],S[33]!==w||S[34]!==D||S[35]!==M||S[36]!==L||S[37]!==B?(k=(0,t.jsx)("div",{className:"max-h-[70dvh] overflow-y-auto",children:(0,t.jsx)(H,{colorByAuthor:w,data:D,dataKey:M,isExpanded:!0,onModelClick:L,valueFormatter:B})}),S[33]=w,S[34]=D,S[35]=M,S[36]=L,S[37]=B,S[38]=k):k=S[38],S[39]!==j||S[40]!==k?(y=(0,t.jsxs)(i.DialogContent,{className:"max-w-3xl",children:[j,k]}),S[39]=j,S[40]=k,S[41]=y):y=S[41],S[42]!==F||S[43]!==y?(T=(0,t.jsx)(i.Dialog,{onOpenChange:R,open:F,children:y}),S[42]=F,S[43]=y,S[44]=T):T=S[44],S[45]!==T||S[46]!==x||S[47]!==p||S[48]!==f?(C=(0,t.jsxs)(n.Card,{className:"p-4",children:[x,p,f,T]}),S[45]=T,S[46]=x,S[47]=p,S[48]=f,S[49]=C):C=S[49],C}function H(e){let r,s,l,o,n,i,c,d,u,m,h,x,g,P=(0,a.c)(35),{colorByAuthor:C,data:S,dataKey:N,isExpanded:w,onModelClick:D,valueFormatter:M}=e,A=void 0!==w&&w,_=A?32:24,B=S.length*_+40;P[0]===Symbol.for("react.memo_cache_sentinel")?(r={bottom:8,left:0,right:8,top:8},s=(0,t.jsx)(v.CartesianGrid,{horizontal:!1,stroke:"hsl(var(--border))",strokeDasharray:"2 4"}),P[0]=r,P[1]=s):(r=P[0],s=P[1]),P[2]===Symbol.for("react.memo_cache_sentinel")?(l={fontSize:11},P[2]=l):l=P[2],P[3]!==M?(o=(0,t.jsx)(y.XAxis,{stroke:"hsl(var(--muted-foreground))",tick:l,tickFormatter:M,type:"number"}),P[3]=M,P[4]=o):o=P[4],P[5]!==A?(n=(0,t.jsx)(O,{isExpanded:A}),P[5]=A,P[6]=n):n=P[6];let K=A?260:144;if(P[7]!==n||P[8]!==K?(i=(0,t.jsx)(T.YAxis,{dataKey:"label",interval:0,stroke:"hsl(var(--muted-foreground))",tick:n,type:"category",width:K}),P[7]=n,P[8]=K,P[9]=i):i=P[9],P[10]!==N||P[11]!==M?(c=(0,t.jsx)($,{dataKey:N,valueFormatter:M}),P[10]=N,P[11]=M,P[12]=c):c=P[12],P[13]===Symbol.for("react.memo_cache_sentinel")?(d={fill:"hsl(var(--muted))"},P[13]=d):d=P[13],P[14]!==c?(u=(0,t.jsx)(k.Tooltip,{content:c,cursor:d,isAnimationActive:!1}),P[14]=c,P[15]=u):u=P[15],P[16]!==C||P[17]!==S||P[18]!==D){let e;P[20]!==C||P[21]!==D?(e=e=>(0,t.jsx)(b.Cell,{cursor:null===e.modelSlug?"default":"pointer",fill:C.get(e.author)??"hsl(var(--muted-foreground))",onClick:null===e.modelSlug?void 0:()=>D(e)},e.id),P[20]=C,P[21]=D,P[22]=e):e=P[22],m=S.map(e),P[16]=C,P[17]=S,P[18]=D,P[19]=m}else m=P[19];return P[23]!==N||P[24]!==m?(h=(0,t.jsx)(p.Bar,{dataKey:N,isAnimationActive:!1,children:m}),P[23]=N,P[24]=m,P[25]=h):h=P[25],P[26]!==S||P[27]!==u||P[28]!==h||P[29]!==o||P[30]!==i?(x=(0,t.jsxs)(f.BarChart,{data:S,layout:"vertical",margin:r,children:[s,o,i,u,h]}),P[26]=S,P[27]=u,P[28]=h,P[29]=o,P[30]=i,P[31]=x):x=P[31],P[32]!==x||P[33]!==B?(g=(0,t.jsx)(j.ResponsiveContainer,{height:B,width:"100%",children:x}),P[32]=x,P[33]=B,P[34]=g):g=P[34],g}function L({data:e,dataKey:t,direction:a}){return e.toSorted((e,r)=>(0,S.compareNullableNumbers)(e[t],r[t],"asc"===a?"ascending":"descending"))}function O(e){let r,s,l,o=(0,a.c)(11),{isExpanded:n,payload:i,x:c,y:d}=e,u=void 0!==n&&n,m=i?.value;if(void 0===m||void 0===c||void 0===d)return null;let h=u?38:20;o[0]!==m||o[1]!==h?(r=m.length>h?`${m.slice(0,h-1).trimEnd()}\u2026`:m,o[0]=m,o[1]=h,o[2]=r):r=o[2];let x=r,g=u?12:10;o[3]!==g||o[4]!==x||o[5]!==c||o[6]!==d?(s=(0,t.jsx)("text",{dominantBaseline:"central",fill:"hsl(var(--muted-foreground))",fontSize:g,textAnchor:"end",x:c,y:d,children:x}),o[3]=g,o[4]=x,o[5]=c,o[6]=d,o[7]=s):s=o[7];let p=s,f=i?.payload?.modelSlug;return o[8]!==p||o[9]!==f?(l=f?(0,t.jsx)("a",{href:`/${f}`,children:p}):p,o[8]=p,o[9]=f,o[10]=l):l=o[10],l}function $(e){let r,s,l=(0,a.c)(6),{active:o,dataKey:n,payload:i,valueFormatter:c}=e,d=i?.[0]?.payload;if(!o||!d)return null;let u=d[n];return l[0]!==u||l[1]!==c?(r=null===u?"--":c(u),l[0]=u,l[1]=c,l[2]=r):r=l[2],l[3]!==d.label||l[4]!==r?(s=(0,t.jsx)(C.BenchmarkChartTooltip,{bodyClassName:"text-sm",label:d.label,children:r}),l[3]=d.label,l[4]=r,l[5]=s):s=l[5],s}function z(e){return{accuracy:t=>(0,N.formatPercent)(e,t/100),cost:N.formatCost.bind(void 0,e),duration:t=>(0,N.formatDurationMs)(e,1e3*t)}}e.s(["Tau2ComparisonCharts",0,function(e){let r,n,i,c,u,p,f,v,b,j,k,y,T,P,C,S,H,O,$,q,V,I,U,G,X,Y,Q,J,W,Z,ee,et,ea,er,es=(0,a.c)(108),{accuracyDescription:el,accuracyTitle:eo,benchmark:en,costDescription:ei,costTitle:ec,durationDescription:ed,durationTitle:eu,metricLabel:em,models:eh}=e,ex=void 0===el?"Representative-run accuracy, best first.":el,eg=void 0===eo?"Accuracy":eo,ep=void 0===en?"tau2-bench-airline":en,ef=void 0===ei?"Average cost per graded task, cheapest first.":ei,ev=void 0===ec?"Cost per task":ec,eb=void 0===ed?"Average wall-clock time per task, fastest first; agents that loop or stall run long.":ed,ej=void 0===eu?"Time per task":eu,ek=void 0===em?"task":em,ey=(0,x.useLocale)(),eT=(0,h.usePosthogClient)();es[0]!==eh?(r=eh.slice(0,10).map(E),es[0]=eh,es[1]=r):r=es[1];let eP=r;es[2]!==eh?(n=()=>eh.slice(0,10).map(K),es[2]=eh,es[3]=n):n=es[3];let[eC,eS]=(0,g.useState)(n);es[4]!==eh?(i=eh.map(B),es[4]=eh,es[5]=i):i=es[5];let eN=i;es[6]!==eC?(c=new Set(eC),es[6]=eC,es[7]=c):c=es[7];let ew=c;if(es[8]!==eh||es[9]!==ew){let e;es[11]!==ew?(e=e=>ew.has(e.modelPermaslug),es[11]=ew,es[12]=e):e=es[12],u=eh.filter(e),es[8]=eh,es[9]=ew,es[10]=u}else u=es[10];let eD=u;es[13]!==eD?(p=eD.map(_),es[13]=eD,es[14]=p):p=es[14];let eM=p;es[15]!==eM?(f=(0,m.createStableColorMapping)([...new Set(eM.map(A))]),es[15]=eM,es[16]=f):f=es[16];let eA=f;es[17]!==eD?(v=eD.toSorted(M)[0]??null,es[17]=eD,es[18]=v):v=es[18];let e_=v;es[19]!==eD?(b=(0,s.findBestValueModel)(eD),es[19]=eD,es[20]=b):b=es[20];let eB=b;es[21]!==eD?(j=eD.toSorted(D).find(w)??null,es[21]=eD,es[22]=j):j=es[22];let eK=j;es[23]!==eP||es[24]!==eC?(k=eC.length===eP.length&&eC.every(e=>eP.includes(e)),es[23]=eP,es[24]=eC,es[25]=k):k=es[25];let eE=k;es[26]!==ep||es[27]!==eT?(y=e=>{eT.capture(l.PostHogEvent.ViewBenchmarkModel,{benchmark:ep,chart:"winner_card",model_slug:e})},es[26]=ep,es[27]=eT,es[28]=y):y=es[28];let eF=y;return 0===eh.length?null:(es[29]===Symbol.for("react.memo_cache_sentinel")?(T=(0,t.jsx)(d.MultiSelectTrigger,{placeholder:"Select models..."}),es[29]=T):T=es[29],es[30]!==ep||es[31]!==eP||es[32]!==eT?(P=()=>{eT.capture(l.PostHogEvent.ResetBenchmarkComparisonModels,{benchmark:ep}),eS(eP)},es[30]=ep,es[31]=eP,es[32]=eT,es[33]=P):P=es[33],es[34]!==eE||es[35]!==P?(C=(0,t.jsx)(d.MultiSelectContent,{footer:(0,t.jsx)(o.Button,{className:"w-full",disabled:eE,onClick:P,variant:"ghost",children:"Reset to top 10"}),searchPlaceholder:"Search models..."}),es[34]=eE,es[35]=P,es[36]=C):C=es[36],es[37]!==eN||es[38]!==eC||es[39]!==C?(S=(0,t.jsx)("div",{className:"w-full sm:max-w-xs",children:(0,t.jsxs)(d.MultiSelect,{onValueChange:eS,options:eN,size:"compact",value:eC,children:[T,C]})}),es[37]=eN,es[38]=eC,es[39]=C,es[40]=S):S=es[40],es[41]!==ey||es[42]!==e_?(H=null===e_?null:(0,N.formatPercent)(ey,e_.median.accuracy),es[41]=ey,es[42]=e_,es[43]=H):H=es[43],es[44]!==eF||es[45]!==e_||es[46]!==H?(O=(0,t.jsx)(F,{label:"Most Accurate",model:e_,value:H,onSelect:eF}),es[44]=eF,es[45]=e_,es[46]=H,es[47]=O):O=es[47],es[48]!==eB||es[49]!==ey||es[50]!==ek?($=null===eB?null:`${(0,N.formatCost)(ey,eB.median.avgCostPerTask)}/${ek}`,es[48]=eB,es[49]=ey,es[50]=ek,es[51]=$):$=es[51],es[52]!==eB||es[53]!==eF||es[54]!==$?(q=(0,t.jsx)(F,{label:"Best Value",model:eB,value:$,onSelect:eF}),es[52]=eB,es[53]=eF,es[54]=$,es[55]=q):q=es[55],es[56]!==eK||es[57]!==ey?(V=null===eK?null:(0,N.formatDurationMs)(ey,eK.median.avgDurationPerTaskMs),es[56]=eK,es[57]=ey,es[58]=V):V=es[58],es[59]!==eK||es[60]!==eF||es[61]!==V?(I=(0,t.jsx)(F,{label:"Fastest",model:eK,value:V,onSelect:eF}),es[59]=eK,es[60]=eF,es[61]=V,es[62]=I):I=es[62],es[63]!==O||es[64]!==q||es[65]!==I?(U=(0,t.jsxs)("div",{className:"grid grid-cols-1 gap-4 lg:grid-cols-3",children:[O,q,I]}),es[63]=O,es[64]=q,es[65]=I,es[66]=U):U=es[66],es[67]!==eM?(G=L({data:eM,dataKey:"accuracyPercent",direction:"desc"}),es[67]=eM,es[68]=G):G=es[68],es[69]!==ey?(X=z(ey),es[69]=ey,es[70]=X):X=es[70],es[71]!==ex||es[72]!==eg||es[73]!==ep||es[74]!==eA||es[75]!==G||es[76]!==X.accuracy?(Y=(0,t.jsx)(R,{benchmark:ep,colorByAuthor:eA,data:G,dataKey:"accuracyPercent",description:ex,title:eg,valueFormatter:X.accuracy}),es[71]=ex,es[72]=eg,es[73]=ep,es[74]=eA,es[75]=G,es[76]=X.accuracy,es[77]=Y):Y=es[77],es[78]!==eM?(Q=L({data:eM,dataKey:"costPerTask",direction:"asc"}),es[78]=eM,es[79]=Q):Q=es[79],es[80]!==ey?(J=z(ey),es[80]=ey,es[81]=J):J=es[81],es[82]!==ep||es[83]!==eA||es[84]!==ef||es[85]!==ev||es[86]!==Q||es[87]!==J.cost?(W=(0,t.jsx)(R,{benchmark:ep,colorByAuthor:eA,data:Q,dataKey:"costPerTask",description:ef,title:ev,valueFormatter:J.cost}),es[82]=ep,es[83]=eA,es[84]=ef,es[85]=ev,es[86]=Q,es[87]=J.cost,es[88]=W):W=es[88],es[89]!==eM?(Z=L({data:eM,dataKey:"secondsPerTask",direction:"asc"}),es[89]=eM,es[90]=Z):Z=es[90],es[91]!==ey?(ee=z(ey),es[91]=ey,es[92]=ee):ee=es[92],es[93]!==ep||es[94]!==eA||es[95]!==eb||es[96]!==ej||es[97]!==Z||es[98]!==ee.duration?(et=(0,t.jsx)(R,{benchmark:ep,colorByAuthor:eA,data:Z,dataKey:"secondsPerTask",description:eb,title:ej,valueFormatter:ee.duration}),es[93]=ep,es[94]=eA,es[95]=eb,es[96]=ej,es[97]=Z,es[98]=ee.duration,es[99]=et):et=es[99],es[100]!==Y||es[101]!==W||es[102]!==et?(ea=(0,t.jsxs)("div",{className:"grid grid-cols-1 gap-4 lg:grid-cols-3",children:[Y,W,et]}),es[100]=Y,es[101]=W,es[102]=et,es[103]=ea):ea=es[103],es[104]!==S||es[105]!==U||es[106]!==ea?(er=(0,t.jsxs)("div",{className:"flex flex-col gap-4",children:[S,U,ea]}),es[104]=S,es[105]=U,es[106]=ea,es[107]=er):er=es[107],er)},"createTau2ComparisonFormatters",0,z])},385361,e=>{"use strict";var t=e.i(222666),a=e.i(505962),r=e.i(553215),s=e.i(956506),l=e.i(126154),o=e.i(20629),n=e.i(252358),i=e.i(779100),c=e.i(61291),d=e.i(190557),u=e.i(270401),m=e.i(4696),h=e.i(249743),x=e.i(73962),g=e.i(952750),p=e.i(907579),f=["axis"],v=(0,u.forwardRef)((e,t)=>u.createElement(p.CartesianChart,{chartName:"ComposedChart",defaultTooltipEventType:"axis",validateTooltipEventTypes:f,tooltipPayloadSearcher:g.arrayTooltipSearcher,categoricalChartProps:e,ref:t})),b=e.i(537457),j=e.i(695942),k=e.i(487154),y=e.i(429088),T=e.i(287172),P=e.i(612960),C=e.i(460867),S=e.i(714170);function N(e){return e.costPerTask}function w(e,t){return e.costPerTask-t.costPerTask}function D(e){return e.isParetoOptimal}function M(e){return e.author}function A(e){return null===e.median.avgCostPerTask?[]:[{accuracyPercent:100*e.median.accuracy,author:e.author,costPerTask:e.median.avgCostPerTask,id:e.modelPermaslug,isParetoOptimal:e.isParetoOptimal,label:e.modelName,modelSlug:e.modelSlug}]}function _(e,t){return t>=1?`${(0,d.formatUsd)(t,{locale:e,preset:"native-fixed-0"})??"—"}`:t>=.1?`${(0,d.formatUsd)(t,{locale:e,preset:"native-fixed-1"})??"—"}`:`${(0,d.formatUsd)(t,{locale:e,preset:"native-fixed-2"})??"—"}`}function B(e){let r,s,o,n,c,d,m,h,x=(0,a.c)(22),{active:g,metricLabel:p,onActivePointChange:f,payload:v}=e,b=(0,i.useLocale)(),j=v?.[0]?.payload;if(x[0]!==g||x[1]!==f||x[2]!==j?(r=()=>(f(g&&void 0!==j?j:null),()=>{f(null)}),s=[g,f,j],x[0]=g,x[1]=f,x[2]=j,x[3]=r,x[4]=s):(r=x[3],s=x[4]),(0,u.useEffect)(r,s),!g||!j)return null;let k=j.label,y=j.accuracyPercent/100;return x[5]!==b||x[6]!==y?(o=(0,S.formatPercent)(b,y),x[5]=b,x[6]=y,x[7]=o):o=x[7],x[8]===Symbol.for("react.memo_cache_sentinel")?(n=(0,t.jsx)(l.DotSeparator,{}),x[8]=n):n=x[8],x[9]!==b||x[10]!==j.costPerTask?(c=(0,S.formatCost)(b,j.costPerTask),x[9]=b,x[10]=j.costPerTask,x[11]=c):c=x[11],x[12]!==p||x[13]!==o||x[14]!==c?(d=(0,t.jsxs)("span",{children:[o," ",n," ",c,"/",p]}),x[12]=p,x[13]=o,x[14]=c,x[15]=d):d=x[15],x[16]!==j.modelSlug?(m=j.modelSlug&&(0,t.jsx)("span",{className:"block text-xs text-muted-foreground",children:"Click to view model"}),x[16]=j.modelSlug,x[17]=m):m=x[17],x[18]!==j.label||x[19]!==d||x[20]!==m?(h=(0,t.jsxs)(C.BenchmarkChartTooltip,{label:k,children:[d,m]}),x[18]=j.label,x[19]=d,x[20]=m,x[21]=h):h=x[21],h}e.s(["Tau2ParetoChart",0,function(e){let l,d,g,p,f,C,S,K,E,F,R,H,L,O,$,z,q,V,I,U,G,X,Y,Q,J,W,Z,ee,et,ea,er,es=(0,a.c)(93),{benchmark:el,metricLabel:eo,models:en}=e,ei=void 0===el?"tau2-bench-airline":el,ec=void 0===eo?"task":eo,ed=(0,i.useLocale)(),eu=(0,n.usePosthogClient)(),em=(0,P.useLocalizedRouter)(),eh=(0,u.useRef)(null);es[0]!==en?(l=en.flatMap(A),es[0]=en,es[1]=l):l=es[1];let ex=l;es[2]!==ex?(d=(0,o.createStableColorMapping)([...new Set(ex.map(M))]),es[2]=ex,es[3]=d):d=es[3];let eg=d;es[4]!==ex?(g=ex.filter(D).toSorted(w),es[4]=ex,es[5]=g):g=es[5];let ep=g;es[6]!==ex?(p=function(e){if(0===e.length)return[];let t=Math.min(...e),a=Math.max(...e);return[1,2,5].flatMap(e=>[-4,-3,-2,-1,0,1,2].map(t=>e*10**t)).filter(e=>e>=t&&e<=a).toSorted((e,t)=>e-t)}(ex.map(N)),es[6]=ex,es[7]=p):p=es[7];let ef=p;es[8]===Symbol.for("react.memo_cache_sentinel")?(f=e=>{eh.current=e},es[8]=f):f=es[8];let ev=f;if(0===ex.length)return null;es[9]!==ei||es[10]!==eu||es[11]!==em?(C=e=>{null!==e.modelSlug&&(eu.capture(r.PostHogEvent.ViewBenchmarkModel,{benchmark:ei,chart:"pareto",model_slug:e.modelSlug}),em.push(`/${e.modelSlug}`))},es[9]=ei,es[10]=eu,es[11]=em,es[12]=C):C=es[12];let eb=C;if(es[13]!==eg||es[14]!==ef||es[15]!==ep||es[16]!==eb||es[17]!==ed||es[18]!==ec||es[19]!==ex){let e,a,r,l,o,n,i;F=s.Card,Y="p-4",es[40]===Symbol.for("react.memo_cache_sentinel")?(Q=(0,t.jsx)(s.CardTitle,{children:"Accuracy vs. cost (Pareto frontier)"}),J=(0,t.jsx)(s.CardDescription,{className:"mt-1",children:"One point per model, using default routing (not pinned to a provider) when available. The line is the Pareto frontier: no model beats these on both accuracy and cost."}),es[40]=Q,es[41]=J):(Q=es[40],J=es[41]),X="mt-4 h-80",E=b.ResponsiveContainer,U="100%",G="100%",K=v,es[42]===Symbol.for("react.memo_cache_sentinel")?(L=()=>ev(null),O={bottom:8,left:0,right:16,top:8},$=(0,t.jsx)(m.CartesianGrid,{stroke:"hsl(var(--border))",strokeDasharray:"2 4"}),es[42]=L,es[43]=O,es[44]=$):(L=es[42],O=es[43],$=es[44]);let d=`Cost per ${ec}`;es[45]===Symbol.for("react.memo_cache_sentinel")?(e=["auto","auto"],es[45]=e):e=es[45],es[46]===Symbol.for("react.memo_cache_sentinel")?(a={fontSize:12},es[46]=a):a=es[46],es[47]!==ed?(r=_.bind(void 0,ed),es[47]=ed,es[48]=r):r=es[48],es[49]!==ef||es[50]!==d||es[51]!==r?(z=(0,t.jsx)(y.XAxis,{dataKey:"costPerTask",name:d,scale:"log",domain:e,stroke:"hsl(var(--muted-foreground))",tick:a,tickFormatter:r,ticks:ef,type:"number"}),es[49]=ef,es[50]=d,es[51]=r,es[52]=z):z=es[52],es[53]===Symbol.for("react.memo_cache_sentinel")?(l=["auto","auto"],es[53]=l):l=es[53],es[54]===Symbol.for("react.memo_cache_sentinel")?(o={fontSize:12},es[54]=o):o=es[54],es[55]!==ed?(q=(0,t.jsx)(T.YAxis,{dataKey:"accuracyPercent",name:"Accuracy",domain:l,stroke:"hsl(var(--muted-foreground))",tick:o,tickFormatter:e=>(0,c.formatPercentPoints)(Math.round(e),{locale:ed,preset:"fee"})??"—",type:"number",width:48}),es[55]=ed,es[56]=q):q=es[56],es[57]!==ec?(n=(0,t.jsx)(B,{metricLabel:ec,onActivePointChange:ev}),es[57]=ec,es[58]=n):n=es[58],es[59]===Symbol.for("react.memo_cache_sentinel")?(i={stroke:"hsl(var(--border))",strokeDasharray:"2 4"},es[59]=i):i=es[59],es[60]!==n?(V=(0,t.jsx)(k.Tooltip,{content:n,cursor:i}),es[60]=n,es[61]=V):V=es[61],es[62]!==ep?(I=(0,t.jsx)(x.Line,{data:ep,dataKey:"accuracyPercent",dot:!1,isAnimationActive:!1,stroke:"hsl(var(--muted-foreground))",strokeDasharray:"4 4",type:"stepAfter"}),es[62]=ep,es[63]=I):I=es[63],S=j.Scatter,W=ex,R=!1,H=ex.map(e=>(0,t.jsx)(h.Cell,{cursor:null===e.modelSlug?"default":"pointer",fill:eg.get(e.author)??"hsl(var(--muted-foreground))",onClick:null===e.modelSlug?void 0:()=>{eb(eh.current??e)}},e.id)),es[13]=eg,es[14]=ef,es[15]=ep,es[16]=eb,es[17]=ed,es[18]=ec,es[19]=ex,es[20]=S,es[21]=K,es[22]=E,es[23]=F,es[24]=R,es[25]=H,es[26]=L,es[27]=O,es[28]=$,es[29]=z,es[30]=q,es[31]=V,es[32]=I,es[33]=U,es[34]=G,es[35]=X,es[36]=Y,es[37]=Q,es[38]=J,es[39]=W}else S=es[20],K=es[21],E=es[22],F=es[23],R=es[24],H=es[25],L=es[26],O=es[27],$=es[28],z=es[29],q=es[30],V=es[31],I=es[32],U=es[33],G=es[34],X=es[35],Y=es[36],Q=es[37],J=es[38],W=es[39];return es[64]!==S||es[65]!==R||es[66]!==H||es[67]!==W?(Z=(0,t.jsx)(S,{data:W,isAnimationActive:R,children:H}),es[64]=S,es[65]=R,es[66]=H,es[67]=W,es[68]=Z):Z=es[68],es[69]!==K||es[70]!==L||es[71]!==O||es[72]!==$||es[73]!==z||es[74]!==q||es[75]!==V||es[76]!==I||es[77]!==Z?(ee=(0,t.jsxs)(K,{onMouseLeave:L,margin:O,children:[$,z,q,V,I,Z]}),es[69]=K,es[70]=L,es[71]=O,es[72]=$,es[73]=z,es[74]=q,es[75]=V,es[76]=I,es[77]=Z,es[78]=ee):ee=es[78],es[79]!==E||es[80]!==U||es[81]!==G||es[82]!==ee?(et=(0,t.jsx)(E,{height:U,width:G,children:ee}),es[79]=E,es[80]=U,es[81]=G,es[82]=ee,es[83]=et):et=es[83],es[84]!==X||es[85]!==et?(ea=(0,t.jsx)("div",{className:X,children:et}),es[84]=X,es[85]=et,es[86]=ea):ea=es[86],es[87]!==F||es[88]!==Y||es[89]!==Q||es[90]!==J||es[91]!==ea?(er=(0,t.jsxs)(F,{className:Y,children:[Q,J,ea]}),es[87]=F,es[88]=Y,es[89]=Q,es[90]=J,es[91]=ea,es[92]=er):er=es[92],er},"formatCostTick",0,_],385361)}]); //# sourceMappingURL=291dez9370b-n.js.map //# chunkId=01a08fa9-635e-7b13-9b2d-0367079a54a2