Rubrica
Aprovação ou reprovação binária por verificação
O modelo seguiu as instruções da tarefa, identificou requisitos ocultos nos arquivos de origem, usou as evidências corretas e chegou às conclusões adequadas?
AA-Briefcase v1.1: Agentic Knowledge Work Benchmark
Índice de InteligênciaConjunto de dados privadoAgênticoTrabalho do conhecimentoUso de ferramentasRedaçãoRaciocínioSeguimento de instruçõesNegócios
A private evaluation developed by Artificial Analysis for frontier agentic capability in long-horizon knowledge work, testing agents on realistic business workflows that require deliverables such as spreadsheets, presentations, and memos.
O AA-Briefcase avalia modelos em quatro projetos de trabalho do conhecimento com várias semanas de duração, que totalizam milhares de arquivos de entrada e 91 tarefas. Nos diferentes cenários, os modelos precisam concluir fluxos de trabalho profissionais realistas em áreas como ciência de dados, gestão de produtos e estratégia corporativa. Cada cenário é um fluxo de várias semanas percorrido pelo agente em sequência, com várias tarefas por semana. Cada tarefa gera um entregável avaliado por uma rubrica de verificações. Embora as tarefas de um cenário compartilhem arquivos e contexto entre as semanas, atualmente os modelos concluem cada tarefa em uma execução independente, sem reutilizar suas entregas anteriores.
Cada tarefa é avaliada por três tipos de verificação:
Qualidade analítica
Comparação em pares
Em comparação com a entrega de outro modelo, qual resultado é mais completo, rigoroso do ponto de vista analítico e bem fundamentado?
Apresentação
Comparação em pares
Em comparação com a entrega de outro modelo, qual tem uma apresentação mais profissional?
Um quinto cenário público foi disponibilizado no Hugging Face para representar a estrutura, a entrega e a avaliação de um cenário. Ele serve apenas como demonstração e não conta para os resultados oficiais do AA-Briefcase.
Resultados
AA-Briefcase Elo
AA-Briefcase v1.1 is an agentic knowledge work benchmark developed by Artificial Analysis. AA-Briefcase Elo is a combined metric that aggregates rubric pass rate, analytical quality Elo and presentation Elo · Higher is better
Not publicly available
AA-Briefcase Elo vs. Cost per Task
AA-Briefcase Elo · Cost per task (USD)
Most attractive quadrant
Pareto line
Custo
AA-Briefcase Cost per Task
Mean cost (USD) per task to run AA-Briefcase, calculated from token usage and model pricing including representative cache hit rates
Not publicly available
Exemplo de tarefa, entregas e avaliação
Explore uma semana representativa do AA-Briefcase no cenário público de diligência devida disponível no Hugging Face. Os resultados e a avaliação exibidos ilustram o que o AA-Briefcase mede. As pontuações são mostradas para um conjunto representativo de modelos. As entregas e os vereditos deste cenário representativo não contribuem para o Elo do AA-Briefcase nem para outras pontuações do benchmark.
Comparações de pontuação
AA-Briefcase Elo vs. Artificial Analysis Intelligence Index
AA-Briefcase Elo · Artificial Analysis Intelligence Index
Most attractive quadrant
Resultados por tipo de arquivo
Desempenho no AA-Briefcase por tipo de arquivo do entregável (Excel, PowerPoint, PDF, Word ou outro).
AA-Briefcase Rubric Pass Rate by File Type (Normalized)
Rubric pass rate by deliverable file type · Scores are normalized per file type across all models tested, where green represents the highest score for that file type and red represents the lowest score for that file type
Uso de tokens
AA-Briefcase Output Tokens per Task
Mean reasoning and answer tokens consumed per AA-Briefcase task
Not publicly available
Velocidade
Time per Task
Wall-clock time (minutes) per task: answer and reasoning generation plus tool execution time · Lower is better
Turnos
Mean Turns per Task
Average number of model turns per AA-Briefcase task · Lower is better
Not publicly available
Uso de ferramentas
Chamadas de ferramentas feitas por cada agente durante o AA-Briefcase: contagens por categoria, média de chamadas por turno e cobertura da exploração das fontes.
AA-Briefcase Tool Calls Breakdown, Avg per Task
Average tool invocations per AA-Briefcase task, bucketed by intent
Not publicly available
Tamanho do modelo (somente modelos de pesos abertos)
AA-Briefcase Elo vs. Total Parameters
AA-Briefcase Elo · Size in parameters (billions) · Open weights models only
Most attractive quadrant
Pontuação vs. data de lançamento
AA-Briefcase Elo vs. Release Date
AA-Briefcase Elo · Model release date
Most attractive region
Classificação
Criador | Nome | Elo | IC | Data de lançamento | |
|---|---|---|---|---|---|
| 1 |  Anthropic | Claude Opus 5.5 (Adaptive Reasoning, Max Effort, Default Fallback) | 1822 | -12 / +12 | set. de 2026 |
| 2 | Anthropic | Claude Sonnet 5.5 (Adaptive Reasoning, Max Effort, Default Fallback) | 1811 | -11 / +11 | set. de 2026 |
| 3 | Anthropic | Claude Opus 5.5 (Adaptive Reasoning, Xhigh Effort, Default Fallback) | 1780 | -11 / +11 | set. de 2026 |
| 4 | Anthropic | Claude Sonnet 5.5 (Adaptive Reasoning, Xhigh Effort, Default Fallback) | 1746 | -10 / +10 | set. de 2026 |
| 5 | Anthropic | Claude Opus 5.5 (Adaptive Reasoning, High Effort, Default Fallback) | 1705 | -10 / +10 | set. de 2026 |
| 6 | Anthropic | Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) | 1678 | -10 / +10 | set. de 2026 |
| 7 | Anthropic | Claude Opus 5 (Adaptive Reasoning, Max Effort) | 1673 | -10 / +11 | jul. de 2026 |
| 8 | Anthropic | Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback) | 1669 | -9 / +10 | set. de 2026 |
| 9 |  SpaceXAI | Grok 4.7 (Xhigh) | 1657 | -9 / +9 | set. de 2026 |
| 10 | Anthropic | Claude Opus 5 (Adaptive Reasoning, Xhigh Effort) | 1649 | -9 / +10 | jul. de 2026 |
| 11 | Anthropic | Claude Opus 5.5 (Adaptive Reasoning, Medium Effort, Default Fallback) | 1642 | -10 / +10 | set. de 2026 |
| 12 | SpaceXAI | Grok 4.7 (High) | 1637 | -10 / +11 | set. de 2026 |
| 13 | Anthropic | Claude Sonnet 5.5 (Adaptive Reasoning, High Effort, Default Fallback) | 1634 | -10 / +10 | set. de 2026 |
| 14 |  Alibaba | Qwen3.8 Max (0902) | 1621 | -10 / +11 | set. de 2026 |
| 15 | Anthropic | Claude Fable 5.1 (Adaptive Reasoning, High Effort, Default Fallback) | 1592 | -9 / +10 | set. de 2026 |
Explorar avaliações
Artificial Analysis Intelligence Index v4.3.2A composite benchmark aggregating ten challenging evaluations to provide a holistic measure of AI capabilities across mathematics, science, coding, and reasoning.
Artificial Analysis Openness IndexA composite measure providing an industry standard to communicate model openness for users and developers.
AA-Briefcase v1.1: Agentic Knowledge Work BenchmarkA private evaluation developed by Artificial Analysis for frontier agentic capability in long-horizon knowledge work, testing agents on realistic business workflows that require deliverables such as spreadsheets, presentations, and memos.
GDPval-AA v2.1 LeaderboardGDPval-AA v2.1 is Artificial Analysis' evaluation framework for OpenAI's GDPval dataset. It tests AI models on real-world tasks across 44 occupations and 9 major industries. Models are given shell access and web browsing capabilities in an agentic loop via Stirrup to solve tasks, with Elo ratings derived from blind pairwise comparisons.
APEX-Agents-AA Benchmark LeaderboardArtificial Analysis' implementation of the APEX-Agents benchmark, testing AI agents on long-horizon, cross-application tasks in professional-services environments with realistic application tooling.
AA-AnalystAgent Benchmark LeaderboardArtificial Analysis' data analysis benchmark, testing AI agents on their ability to work with spreadsheets and documents to answer quantitative questions a Business Analyst or Data Analyst would face day-to-day.
AutomationBench-AA: Agentic SaaS Workflow BenchmarkA benchmark measuring agentic task completion across simulated SaaS application environments, scoring the share of each task's objectives completed without guardrail violations.
Harvey LAB-AA Benchmark LeaderboardArtificial Analysis' implementation of Harvey's Legal Agent Benchmark (LAB), testing AI agents on real-world legal work from Harvey's dataset of 120 private tasks spanning 24 legal practice areas. The agent reads case documents in a sandbox and produces legal deliverables (e.g., memos, disclosure schedules, deposition summaries), graded criterion-by-criterion by a single LLM rubric judge.
GDP.pdf Benchmark LeaderboardArtificial Analysis' implementation of Surge AI's GDP.pdf benchmark, testing whether language models can reason over long, real-world professional documents and satisfy detailed task-specific criteria.
EnterpriseOps-Gym-AA Benchmark LeaderboardArtificial Analysis' independent implementation of ServiceNow's EnterpriseOps-Gym, an agentic benchmark testing whether LLM agents can complete stateful, multi-step enterprise workflows across eight business domains via live tool use, graded on the final state of the underlying databases.
Terminal-Bench 4.0 Benchmark LeaderboardA harder 66-task benchmark of complex terminal work across software, machine learning, science, operations, security, hardware, and media, with recalibrated compute and time allowances and improved instructions, environments, and verifiers.
Terminal-Bench-Science 0.1 Benchmark LeaderboardA 70-task benchmark of research workflows authored and reviewed by domain experts across the life, physical, earth, mathematical, and engineering sciences, each completed in a terminal and checked by its own set of tests.
Artificial Analysis Long Context Reasoning Benchmark LeaderboardA challenging benchmark measuring language models' ability to extract, reason about, and synthesize information from long-form documents ranging from 10k to 100k tokens (measured using the cl100k_base tokenizer).
AA-Omniscience: Knowledge and Hallucination BenchmarkA benchmark measuring factual recall and hallucination across various economically relevant domains.
SciCode Benchmark LeaderboardA scientist-curated coding benchmark featuring 288 test set subproblems from 80 laboratory problems across 16 scientific disciplines.
Humanity's Last Exam Benchmark LeaderboardA frontier-level benchmark with 2,500 expert-vetted questions across mathematics, sciences, and humanities, designed to be the final closed-ended academic evaluation.
CritPt Benchmark LeaderboardA benchmark designed to test LLMs on research-level physics reasoning tasks, featuring 71 composite research challenges.
GPQA Diamond Benchmark Leaderboard
The most challenging 198 questions from GPQA, where PhD experts achieve 65% accuracy but skilled non-experts only reach 34% despite web access.
ITBench-AA Benchmark LeaderboardArtificial Analysis' implementation of IBM's ITBench benchmark, testing AI agents on Kubernetes incident root-cause analysis from offline incident snapshots. The agent inspects alerts, events, traces, and topology and identifies the contributing-factor entities (deployments, pods, namespaces, network policies, etc.) responsible for the failure.
MMMU-Pro Benchmark LeaderboardAn enhanced MMMU benchmark that eliminates shortcuts and guessing strategies to more rigorously test multimodal models across 30 academic disciplines.
IFBench Benchmark LeaderboardA benchmark evaluating precise instruction-following generalization on 58 diverse, verifiable out-of-domain constraints that test models' ability to follow specific output requirements.
Medical Long Context Reasoning (MLCR-AA)An open benchmark from Wisedocs measuring how well models reason over long, fragmented medical records, performing the multi-document synthesis claims professionals rely on when reviewing insurance and healthcare cases.
𝜏³-Banking Benchmark LeaderboardA fintech customer-support benchmark from the 𝜏-Knowledge framework that tests whether agents can navigate a large unstructured knowledge base and execute multi-step tool calls to resolve realistic banking workflows.
Terminal-Bench 2.1 Benchmark LeaderboardA verified refresh of Terminal-Bench 2.0 — 89 curated tasks across software engineering, system administration, data processing, model training, and security, with environment and instruction fixes so scores reflect agent capability rather than environment gaps.
Terminal-Bench Hard Benchmark LeaderboardAn agentic benchmark evaluating AI capabilities in terminal environments through software engineering, system administration, and data processing tasks.
𝜏²-Bench Telecom Benchmark LeaderboardA dual-control conversational AI benchmark simulating technical support scenarios where both agent and user must coordinate actions to resolve telecom service issues.
MMLU-Pro Benchmark LeaderboardAn enhanced version of MMLU with 12,000 graduate-level questions across 14 subject areas, featuring ten answer options and deeper reasoning requirements.
LiveCodeBench Benchmark LeaderboardA contamination-free coding benchmark that continuously harvests fresh competitive programming problems from LeetCode, AtCoder, and CodeForces, evaluating code generation, self-repair, and execution.
MATH-500 Benchmark LeaderboardA 500-problem subset from the MATH dataset, featuring competition-level mathematics across six domains including algebra, geometry, and number theory.
AIME 2025 Benchmark LeaderboardAll 30 problems from the 2025 American Invitational Mathematics Examination, testing olympiad-level mathematical reasoning with integer answers from 000-999.
Global-MMLU-Lite Benchmark LeaderboardA lightweight, multilingual version of MMLU, designed to evaluate knowledge and reasoning skills across a diverse range of languages and cultural contexts.