PhaseoPhaseo
ModelosChatCompararAplicativosClassificações

Ajude a manter os dados públicos corretos

Viu algo faltando, incorreto ou desatualizado? Conte-nos sobre esta página.

PhaseoPhaseo
Verificando statusVerificando statusVisitar página de status
O status por componente está indisponível.

Explorar

  • Modelos
  • Chat
  • Provedores
  • Aplicativos
  • Classificações
  • Ferramentas

Recursos

  • Comparar
  • Guias de migração
  • Metodologia
  • Blog

Comunidade

  • Discord
  • GitHub
  • LinkedIn
  • Reddit
  • X

Criar

  • Documentação
  • Referência da API
  • Início rápido
  • SDKs

Empresa

  • Sobre
  • Central de confiança
  • Missão
  • Preços
  • Funciona com
  • Agradecimentos
  • Suporte
  • Privacidade
  • Termos

Explorar

  • Modelos
  • Chat
  • Provedores
  • Aplicativos
  • Classificações
  • Ferramentas

Criar

  • Documentação
  • Referência da API
  • Início rápido
  • SDKs

Recursos

  • Comparar
  • Guias de migração
  • Metodologia
  • Blog

Empresa

  • Sobre
  • Central de confiança
  • Missão
  • Preços
  • Funciona com
  • Agradecimentos
  • Suporte
  • Privacidade
  • Termos

Comunidade

  • Discord
  • GitHub
  • LinkedIn
  • Reddit
  • X

© 2025 • Phaseo

Alterar idioma: Português (Brasil)
  • English (UK)
  • 简体中文
  • हिन्दी
  • Español (España)
  • Français (France)
  • Deutsch (Deutschland)
  • Português (Brasil)
  • 日本語
  • العربية

Ajuda:problema·Suporte

Precisa de ajuda com o Phaseo?Abrir um problemaouentrar em contato com o suporte

Criar conta

Artificial Analysis

Relatar um problema
v5.0v4

Custo e inteligência

Tempo por tarefa e inteligência

Tokens de saída por tarefa

LBaixoMMédioHAltoXMuito altoMxMáximo

Total de tokens de saída

LBaixoMMédioHAltoXMuito altoMxMáximo

Custo por tarefa

LBaixoMMédioHAltoXMuito altoMxMáximo

Custo total da avaliação

LBaixoMMédioHAltoXMuito altoMxMáximo

Benchmarks individuais

Configurações avaliadas
Modelo / configuração
Claude Opus 5.5 (Low, Default Fallback)0,3130,5860,483—39,0330,8071.240,150,1770,2560,5290,847———————
Claude Opus 5.5 (Medium, Default Fallback)0,5250,5930,547—39,7670,8431.592,560,2770,2560,6120,857———————
Claude Opus 5.5 (High, Default Fallback)0,5660,6040,556—42,3670,8271.713,60,3090,2880,6320,858———————
Claude Opus 5.5 (Xhigh, Default Fallback)0,5960,650,575—43,1330,8471.835,030,3170,2660,650,866———————
Claude Opus 5.5 (Max, Default Fallback)0,5960,6690,614—48,1330,8471.866,270,3170,2620,6950,877———————
GPT-6 Astra (Low)0,4190,5410,4920,93140,30,81.403,260,2630,3040,5910,846————0,32—0,88
GPT-6 Astra (Medium)0,4950,5420,5270,93941,40,7971.498,470,2910,3040,6460,851————0,355—0,895
GPT-6 Astra (High)0,540,5540,5310,94944,5670,81.536,040,2890,310,6660,864————0,4—0,899
GPT-6 Astra (Xhigh)0,5960,5570,5460,96343,6670,81.569,360,3140,3220,6720,862————0,431—0,891
GPT-6 Astra (Max)0,5910,5650,5470,96142,4330,8071.583,690,3170,310,6850,869————0,414—0,884
GPT-6.1 Sol (Low)0,3080,5320,474—380,841.321,980,2490,270,5260,831———————
GPT-6.1 Sol (Medium)0,480,5320,499—40,1670,8331.459,050,2770,30,6260,839———————
GPT-6.1 Sol (High)0,5150,5580,514—410,8231.520,450,30,320,6450,849———————
GPT-6.1 Sol (Xhigh)0,540,5570,526—42,3670,7971.554,80,3170,3180,6660,851———————
GPT-6.1 Sol (Max)0,5610,5420,529—42,0330,831.608,890,3170,310,6490,86———————
GPT-6.1 Terra (Low)0,1520,5240,358—22,9330,811.203,750,1510,2560,5620,825———————
GPT-6.1 Terra (Medium)0,3230,5160,402—26,60,8071.324,70,2140,2840,5780,828———————
GPT-6.1 Terra (High)0,4190,5160,433—27,7670,8231.352,740,2570,280,5850,838———————
GPT-6.1 Terra (Xhigh)0,4650,5340,437—28,9330,8171.414,870,2660,2660,5990,838———————
GPT-6.1 Terra (Max)0,5560,5510,462—29,5670,8271.475,630,280,2860,6020,847———————
Gemini 4 Argon (High)0,5710,6180,571—46,10,7971.630,660,2710,2180,775————————
Muse Wonder Full (Max)0,5710,6260,536—30,60,8371.857,090,3290,3260,713————————
Claude Fable 5.1 (Low, Default Fallback)0,4040,5670,4890,88133,9330,8231.476,70,2770,280,522—————0,39—0,85
Claude Fable 5.1 (Medium, Default Fallback)0,4490,5640,5380,88636,4330,8471.554,230,2910,2680,547—————0,41—0,88
Claude Fable 5.1 (High, Default Fallback)0,520,5870,5590,90639,9330,8371.644,080,3030,2680,553—————0,431—0,899
Claude Fable 5.1 (Xhigh, Default Fallback)0,5510,6090,5870,93444,3330,831.740,220,3110,2620,578—————0,458—0,91
Claude Fable 5.1 (Max, Default Fallback)0,520,6310,5910,93746,2670,8531.763,290,2970,2620,594—————0,472—0,914
Muse Wonder (Max)0,520,6030,501—25,10,8071.770,340,2830,3060,681————————
Claude Haiku 5.5 (Low)0,1260,4920,27—2,1670,721.130,360,0910,1120,229————————
Claude Haiku 5.5 (Medium)0,1520,490,338—4,1670,7731.282,50,1290,1520,286————————
Claude Haiku 5.5 (High)0,2170,4870,373—5,30,7731.424,290,1860,1720,337————————
Claude Haiku 5.5 (Xhigh)0,2930,5170,427—7,8330,7831.521,540,2260,1820,36————————
Claude Haiku 5.5 (Max)0,3280,550,444—12,30,8271.630,830,1890,2080,354————————
Grok 4.7 (Low)0,1620,5490,392—27,40,81.591,40,1340,2140,5670,782———————
Grok 4.7 (High)0,2470,5780,423—300,771.715,80,180,2320,635————————
Grok 4.7 (Xhigh)0,2580,5740,431—31,2330,7671.718,420,1770,20,656————————
GPT-6 Luna (Low)00,4690,203—-18,10,741.040,480,0260,090,120,76———————
GPT-6 Luna (Medium)0,0250,5090,283—-13,5670,7831.270,450,1060,1380,4050,779———————
GPT-6 Luna (High)0,0450,5030,329—-12,80,7931.352,10,1540,1820,4780,775———————
GPT-6 Luna (Xhigh)0,0810,5170,343—-12,9670,81.372,520,1740,1860,4780,797———————
GPT-6 Luna (Max)0,1260,5460,385—-8,2670,8331.444,460,1940,2280,5320,797———————

Capacidades por área

Configurações avaliadas
Modelo / configuração
Claude Opus 5.5 (Low, Default Fallback)42,17447,50443,47645,1442,96745,155
Claude Opus 5.5 (Medium, Default Fallback)50,74555,98849,7151,4253,59954,329
Claude Opus 5.5 (High, Default Fallback)53,25557,56951,95352,48357,69656,892
Claude Opus 5.5 (Xhigh, Default Fallback)55,7259,00254,72355,51462,47260,004
Claude Opus 5.5 (Max, Default Fallback)59,39961,44358,03957,91764,57662,327
GPT-6 Astra (Low)48,12149,87947,68844,98249,52447,826
GPT-6 Astra (Medium)53,69155,66351,82949,6456,69853,901
GPT-6 Astra (High)56,13457,19655,06251,56559,97756,645
GPT-6 Astra (Xhigh)56,8157,86155,65452,6560,32957,793
GPT-6 Astra (Max)57,37857,97555,56953,73161,79558,322
GPT-6.1 Sol (Low)44,2545,57342,92739,67842,2743,438
GPT-6.1 Sol (Medium)50,60253,25949,00245,81751,85350,267
GPT-6.1 Sol (High)53,17855,44351,77348,81355,27653,669
GPT-6.1 Sol (Xhigh)55,20957,23853,68250,75657,30155,716
GPT-6.1 Sol (Max)56,22857,61353,66452,03459,25357,424
GPT-6.1 Terra (Low)37,94342,17637,256—35,01635,996
GPT-6.1 Terra (Medium)42,49545,44940,499—42,56241,067
GPT-6.1 Terra (High)44,20847,44942,5—46,32543,627
GPT-6.1 Terra (Xhigh)47,63648,65943,765—49,11746,741
GPT-6.1 Terra (Max)48,99450,01546,38444,72354,38649,323
Gemini 4 Argon (High)55,92457,20154,57651,28257,30253,963
Muse Wonder Full (Max)55,22857,89850,72551,88654,92155,565
Claude Fable 5.1 (Low, Default Fallback)47,71951,13446,68844,06445,36749,178
Claude Fable 5.1 (Medium, Default Fallback)47,46952,30548,08245,31150,04850,997
Claude Fable 5.1 (High, Default Fallback)49,61353,79349,81747,8954,14853,727
Claude Fable 5.1 (Xhigh, Default Fallback)53,50956,43553,73150,53756,88457,944
Claude Fable 5.1 (Max, Default Fallback)53,8257,96754,49751,17456,58958,933
Muse Wonder (Max)51,1354,48246,86948,70349,30350,886
Claude Haiku 5.5 (Low)28,50727,28827,11824,83926,84729,548
Claude Haiku 5.5 (Medium)34,87732,97332,60430,69831,79535,55
Claude Haiku 5.5 (High)37,84337,1135,03932,19136,25339,44
Claude Haiku 5.5 (Xhigh)41,10739,13837,14734,11340,2642,633
Claude Haiku 5.5 (Max)43,55939,11538,66737,49842,28844,348
Grok 4.7 (Low)45,29949,25844,40242,89737,43742,752
Grok 4.7 (High)50,62153,1546,62346,35641,46647,238
Grok 4.7 (Xhigh)51,68954,82747,19646,59940,95147,584
GPT-6 Luna (Low)17,17516,80116,34816,16818,72718,722
GPT-6 Luna (Medium)26,28533,75824,45927,40424,77727,503
GPT-6 Luna (High)29,2637,78826,66329,24827,20729,847
GPT-6 Luna (Xhigh)30,55139,39328,38230,48928,83932,105
GPT-6 Luna (Max)33,86843,18832,21134,82433,19935,485
Modelos 11
/

Raciocínio

Avançados
Destaques
Fonte: Artificial Analysis
Índice v5.0Última sincronização 31 de out. de 2026
Configurações avaliadas41
Configurações avaliadas
Modelo / configuraçãoInteligênciaTokens de saída por tarefaTotal de tokens de saídaCusto por tarefa (USD)Custo total da avaliação (USD)Tempo por tarefa (s)
Claude Fable 5.1 (Low, Default Fallback)38,483648810737.836,512721236.344.4454,471135894653.885,93502864413,730600196
GPT-6.1 Sol (Xhigh)52,308219253330.241,183947533.286.2830,763994935616717,708156949425,698330478
GPT-6 Luna (Xhigh)23,300872808943.893,214103356.118.4510,10409320921185,744011654362,243332216
GPT-6 Astra (Xhigh)55,25574685726.723,758595432.284.6723,658926490613.746,21535857406,085509866
Claude Haiku 5.5 (High)29,624315467788.064,347804989.814.0620,158560319567121,413039537305,719268626
Grok 4.7 (Low)30,882077602642.669,584907651.253.4682,922201301192.168,64512758588,121973091
Grok 4.7 (High)35,01229629689.127,175774123.196.0746,183595565384.721,607424971.113,10958533
GPT-6 Astra (Max)56,853855959751.973,198976454.327.8596,24466694365.661,17948256658,382490935
Claude Opus 5.5 (Max, Default Fallback)58,4060336079173.312,638857218.145.40810,03640971958.909,667015651.140,77864281
GPT-6 Luna (Medium)18,304215619219.823,2381824.887.3510,042201836106539,5503276316—
Claude Fable 5.1 (Medium, Default Fallback)42,971597828348.545,495652449.158.5315,564264690224.896,57665942536,274658397
GPT-6.1 Sol (High)50,229123193722.606,199071824.651.1300,590985351578573,167791013374,606289867
Claude Fable 5.1 (Max, Default Fallback)49,0102124158117.940,937972163.073.49712,993997201213.326,77852231.073,88598588
Claude Haiku 5.5 (Xhigh)34,3387759487136.176,990729150.911.6150,248361101266192,695109987466,539286921
Claude Haiku 5.5 (Low)18,618005531128.160,550009229.899.4880,04375768638239,6028900779115,218049314
Claude Opus 5.5 (Low, Default Fallback)34,502785007819.059,557465721.114.6340,975529322446979,656483804153,000427601
Grok 4.7 (Xhigh)34,335011706792.969,506435142.604.7457,511769586415.784,27246391.089,75456883
GPT-6 Luna (Max)27,755200493875.296,0523656109.588.3160,151955757321136,060821851631,021642241
Claude Haiku 5.5 (Max)36,234543588225.965,959716297.721.9150,411007415368339,472802241592,373239664
GPT-6 Luna (High)21,379017086532.839,141229240.451.6070,072947764731962,5750472458287,856450326
Claude Fable 5.1 (Xhigh, Default Fallback)49,671638424997.058,7880449114.929.20610,847474706110.096,622785970,62018404
Claude Fable 5.1 (High, Default Fallback)47,024988245666.030,713467269.641.4427,539956160236.617,89007374733,231047376
GPT-6.1 Sol (Medium)46,146879706713.518,335935215.120.8330,359389864803388,568776851264,321113566
GPT-6 Astra (Low)43,03498232299.125,8089996110.270.3071,423722963571.650,74841495156,962743118
Claude Haiku 5.5 (Medium)24,551325434353.335,116519852.730.4060,091764558406772,6994579627195,301464617
GPT-6.1 Terra (Max)50,976518857759.883,164717999.386.0520,961108031082972,335839003610,145274863
GPT-6.1 Sol (Max)54,508733813867.823,691737863.661.0061,272063267941.123,97392095927,671787286
GPT-6 Luna (Low)9,666108123363.295,890335395.823.0650,005115255188529,0871927933526,2601585366
GPT-6.1 Terra (Medium)38,115207667810.875,462740813.627.2170,168804582384190,360237377118,859280809
GPT-6.1 Terra (Low)29,21692807924.422,172267517.189.0280,0713375620517111,16060431248,2504205658
Gemini 4 Argon (High)50,4121935905116.368,029068113.329.2247,682671372526.102,837003631.363,84651098
Claude Opus 5.5 (Xhigh, Default Fallback)56,9947060171103.395,266992100.823.9076,047194490854.765,90876597792,738520875
GPT-6 Astra (High)54,849838322523.133,469636525.584.8563,238571968553.203,9990981356,028422567
GPT-6.1 Sol (Low)34,50292443155.837,994676268.177.4810,161473578318237,907998542112,769004749
GPT-6.1 Terra (Xhigh)44,950051789126.199,659183234.722.7630,400689001667400,175781606283,216645159
GPT-6 Astra (Medium)51,454839386418.051,624427118.172.7042,732438933012.638,28926372318,188702459
Muse Wonder Full (Max)49,727212362788.475,6149841111.539.103———
GPT-6.1 Terra (High)42,334450871316.746,477322420.580.0340,259279646617266,794983365180,100229908
Claude Opus 5.5 (Medium, Default Fallback)48,058102496143.816,546478842.531.1632,451089437662.031,26960729346,911299325
Claude Opus 5.5 (High, Default Fallback)51,685995822760.070,958555456.144.3693,46466228212.715,98000128473,897676499
Muse Wonder (Max)43,324822599896.063,0990918113.389.7086,952036478615.414,45363704682,093132618
Comparações do índice
Evolução

Evolução das pontuações

Compare as pontuações dos modelos por data de lançamento.

Exibindo 86 de 86

Fronteira de Pareto
Modelos com resultados

Resultados de modelos

ClassificaçãoModeloPontuaçãoLançadoFonte
1
Claude Opus 5.5 (Máximo)Anthropic
58,422 de set. de 2026
2
Claude Opus 5.5 (Muito alto)Anthropic
5722 de set. de 2026
3
GPT-6 Astra (Máximo)OpenAI
56,903 de set. de 2026
4
GPT-6 Astra (Muito alto)OpenAI
55,303 de set. de 2026
5
GPT-6 Astra (Alto)OpenAI
54,803 de set. de 2026
6
GPT-6.1 Sol (Máximo)OpenAI
54,529 de set. de 2026
7
Claude Sonnet 5.5 (Máximo)Anthropic
54,228 de set. de 2026
8
GPT-6.1 Sol (Muito alto)OpenAI
52,329 de set. de 2026
9
Claude Opus 5.5 (Alto)Anthropic
51,722 de set. de 2026
10
GPT-6 Astra (Médio)OpenAI
51,503 de set. de 2026
11
GPT-6.1 Terra (Máximo)OpenAI
5126 de out. de 2026
12
Gemini 4 Argon (Alto)Google
50,413 de out. de 2026
13
GPT-6.1 Sol (Alto)OpenAI
50,229 de set. de 2026
14
Claude Fable 5.1 (Muito alto)Anthropic
49,701 de set. de 2026
14
Muse Wonder Full (Máximo)Meta
49,712 de out. de 2026
16
Claude Sonnet 5.5 (Muito alto)Anthropic
49,628 de set. de 2026
17
Claude Fable 5.1 (Máximo)Anthropic
4901 de set. de 2026
18
Claude Opus 5.5 (Médio)Anthropic
48,122 de set. de 2026
19
Claude Fable 5.1 (Alto)Anthropic
4701 de set. de 2026
20
GPT-6.1 Sol (Médio)OpenAI
46,129 de set. de 2026
21
Kimi K3.1 Pro (Padrão)Moonshot
4623 de out. de 2026
22
GPT-6.1 Terra (Muito alto)OpenAI
4526 de out. de 2026
23
Muse Wonder 1.0 (Máximo)Meta
43,312 de out. de 2026
24
GPT-6 Astra (Baixo)OpenAI
4303 de set. de 2026
24
Claude Fable 5.1 (Médio)Anthropic
4301 de set. de 2026
Outros benchmarks7
  • Artificial Analysis Progress Index v1
  • AA-LCR
  • AA-Omniscience
  • AA Omniscience (index) - public split only (our results)
  • AA-Briefcase
  • AA-Index
  • Artificial Analysis Text-to-Video Rank