AA-PersonalityBench: AI Personality and Communication Style Benchmark

Artificial Analysis published AA-PersonalityBench, a benchmark that scores AI models on personality and communication style across 8 indexes built from 30 judged traits and 900 questions, with 30 questions per trait. Each index z-scores its traits' Elo ratings, inverts traits where needed, takes the mean, and percentile ranks the result from 1 to 100 across all evaluated models, with 95% confidence intervals shown per model. The benchmark also groups models into emergent personality families via average-linkage clustering and lets users weight any of the 30 traits from −2 to +2 for a custom ranking.

AA-PersonalityBench: AI Personality & Communication Style Benchmark See example tasks example-tasks AA-PersonalityBench methodology https://artificialanalysis.ai/methodology/aa-personalitybench . Personality Indexes Each index takes the mean of its traits’ Elo z-scores, then percentile ranks it from 1 to 100 across all models evaluated in the benchmark. Personality Indexes How Traits Form Personality Indexes 30 questions per personality trait 900 questions total , judged by a panel of models. Each index z-scores its traits’ Elo ratings, inverts traits where needed, calculates the mean, and percentile ranks the result from 1 to 100 across all models evaluated in the benchmark. Expressiveness150 questionsRich, varied, well-crafted languageLexically VariedEloquentLiterateNatural SoundingMetaphorical https://artificialanalysis.ai/evaluations/aa-personalitybench/indexes/expressiveness Organization120 questionsWell-structured, scannable, answer-firstStructuredScannableBottom-Line SummaryAnswer First https://artificialanalysis.ai/evaluations/aa-personalitybench/indexes/organization Informativeness150 questionsPractical depth, evidence, actionable contentInformation DenseEvidence-BasedResource-SpecificActionable GuidancePragmatic https://artificialanalysis.ai/evaluations/aa-personalitybench/indexes/informativeness Transparency150 questionsTruthful, acknowledges uncertainty, non-sycophanticTruthfulAdmits UncertaintyDisclaims AssumptionsFlags NuancesSycophantic https://artificialanalysis.ai/evaluations/aa-personalitybench/indexes/transparency Warmth120 questionsFriendly, humorous, sincere, emotionally attunedFriendlyHumorousSincereEmotionally Intelligent https://artificialanalysis.ai/evaluations/aa-personalitybench/indexes/warmth Conviction120 questionsDirect, confident positions and constructive pushbackDirectConfidentOpinionatedConstructive Pushback https://artificialanalysis.ai/evaluations/aa-personalitybench/indexes/conviction Imagination60 questionsCreative, exploratory, open-ended thinkingCreativeOpen-Ended https://artificialanalysis.ai/evaluations/aa-personalitybench/indexes/imagination Slop30 questionsUse of filler and generic languageSlop https://artificialanalysis.ai/evaluations/aa-personalitybench/indexes/slop View all indexes https://artificialanalysis.ai/evaluations/aa-personalitybench/indexes Example Questions, Responses, and Grading What's the weirdest social norm we all just accept? Grading summary Answer A leans heavily into observational comedy and playful hyperbole. Its vivid, absurd descriptions like the 'alien perspective' on the birthday cake ritual and the lawn maintenance breakdown make it significantly funnier and more entertaining than Answer B's brief list. Answers were blinded: Answer A was Gemini 4 Argon High and Answer B was MiMo-V2.6-Pro. Detailed Personality Trait Results All 30 judged personality traits which comprise the 8 indexes. Expressiveness Traits https://artificialanalysis.ai/evaluations/aa-personalitybench/indexes/expressiveness Personality Trait Leaderboards Each trait’s Elo percentile ranked from 1 to 100 across all models evaluated in the benchmark. Overall tabs show the index: the mean of its traits’ Elo z-scores, percentile ranked the same way. Whiskers show each model’s 95% confidence interval on the same scale. AA-PersonalityBench: Expressiveness Index https://artificialanalysis.ai/evaluations/aa-personalitybench/indexes/expressiveness Compare Model Personalities Compare any two models: each bar shows the probability one model’s answer is judged to show more of that trait, based on their Elos. AA-PersonalityBench: Expressiveness Index, Gemini 4 Argon high vs GLM-5.3 max Emergent Model Families & Clusters Models grouped by how alike their judged traits and communication characteristics are, using average-linkage clustering on the correlation between z-scores. Emergent Personality Families Build Your Own Personality Ranking Pick any of the 30 traits and weight them from −2 to +2. Your traits’ Elo z-scores are combined as a weighted mean, then percentile ranked from 1 to 100 across all models evaluated in the benchmark. Custom Weighted Personality Ranking Communication Characteristics 21 metrics measured directly from each response, separate from the judged traits above. Cells show the average per response, colored by its percentile rank across all models evaluated in the benchmark. Communication Characteristics: Length Model Fingerprints Each model’s strongest and weakest signals, as standard deviations σ above + or below − the average across models. Model Fingerprints: Personality Traits Claude Opus 5.5 Max, Default Fallback https://artificialanalysis.ai/models/claude-opus-5-5 1. Evidence-Based+2.2σ 2. Information Dense+1.7σ 3. Truthful+1.7σ 4. Literate+1.3σ 5. Lexically Varied+1.3σ 1. Slop−1.0σ 2. Sycophantic−0.2σ 3. Bottom-Line Summary−0.2σ Claude Fable 5.1 Max, Default Fallback https://artificialanalysis.ai/models/claude-fable-5-1 1. Evidence-Based+2.3σ 2. Sincere+1.9σ 3. Opinionated+1.9σ 4. Information Dense+1.9σ 5. Natural Sounding+1.8σ 1. Sycophantic−1.6σ 2. Slop−0.9σ 3. Scannable−0.3σ 4. Answer First+0.0σ GPT-6 Astra Max https://artificialanalysis.ai/models/gpt-6-astra 1. Truthful+1.6σ 2. Admits Uncertainty+1.5σ 3. Disclaims Assumptions+1.4σ 4. Structured+1.3σ 5. Literate+1.0σ 1. Resource-Specific−1.8σ 2. Slop−1.2σ 3. Metaphorical−1.0σ 4. Information Dense−0.6σ 5. Scannable−0.5σ Gemini 4 Argon High https://artificialanalysis.ai/models/gemini-4-argon 1. Metaphorical+2.0σ 2. Humorous+1.9σ 3. Confident+1.6σ 4. Creative+1.1σ 5. Sycophantic+1.1σ 1. Admits Uncertainty−2.1σ 2. Open-Ended−1.5σ 3. Direct−1.3σ 4. Answer First−0.9σ 5. Sincere−0.6σ Muse Wonder Max https://artificialanalysis.ai/models/muse-wonder 1. Friendly+1.8σ 2. Emotionally Intelligent+1.6σ 3. Actionable Guidance+1.4σ 4. Pragmatic+1.3σ 5. Evidence-Based+1.3σ 1. Direct−0.7σ 2. Literate−0.1σ GPT-6.1 Sol Max https://artificialanalysis.ai/models/gpt-6-1-sol 1. Disclaims Assumptions+1.6σ 2. Admits Uncertainty+1.5σ 3. Truthful+1.5σ 4. Structured+1.2σ 5. Direct+1.2σ 1. Resource-Specific−1.8σ 2. Slop−1.2σ 3. Metaphorical−1.1σ 4. Information Dense−0.6σ 5. Actionable Guidance−0.5σ Grok 4.7 Xhigh https://artificialanalysis.ai/models/grok-4-7 1. Direct+1.7σ 2. Sincere+1.2σ 3. Literate+1.1σ 4. Answer First+1.1σ 5. Natural Sounding+0.8σ 1. Sycophantic−1.8σ 2. Slop−1.5σ 3. Friendly−1.3σ 4. Scannable−1.0σ 5. Resource-Specific−1.0σ MiMo-V2.6-Pro https://artificialanalysis.ai/models/mimo-v2-6-pro 1. Scannable+1.6σ 2. Open-Ended+1.1σ 3. Admits Uncertainty+0.5σ 4. Constructive Pushback+0.5σ 5. Slop+0.4σ 1. Answer First−1.5σ 2. Opinionated−1.2σ 3. Actionable Guidance−1.1σ 4. Confident−0.9σ 5. Sycophantic−0.9σ DeepSeek V4.1 Flash 1017 Max https://artificialanalysis.ai/models/deepseek-v4-1-flash-1017 1. Bottom-Line Summary+1.4σ 2. Admits Uncertainty+1.2σ 3. Confident+1.1σ 4. Flags Nuances+1.0σ 5. Actionable Guidance+0.9σ 1. Friendly−1.4σ 2. Sycophantic−1.0σ 3. Emotionally Intelligent−0.8σ 4. Open-Ended−0.8σ 5. Metaphorical−0.5σ GLM-5.3 Max https://artificialanalysis.ai/models/glm-5-3 1. Natural Sounding+1.9σ 2. Lexically Varied+1.5σ 3. Eloquent+1.5σ 4. Sincere+1.5σ 5. Constructive Pushback+1.3σ 1. Actionable Guidance−0.9σ 2. Slop−0.7σ 3. Sycophantic−0.7σ 4. Resource-Specific−0.6σ 5. Bottom-Line Summary−0.5σ Explore Evaluations Artificial Analysis Intelligence Index v4.3.2 https://artificialanalysis.ai/evaluations/artificial-analysis-intelligence-index A composite benchmark aggregating ten challenging evaluations to provide a holistic measure of AI capabilities across mathematics, science, coding, and reasoning. AA-Briefcase v1.1: Agentic Knowledge Work Benchmark https://artificialanalysis.ai/evaluations/aa-briefcase 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. GDPval-AA v2.1 Benchmark https://artificialanalysis.ai/evaluations/gdpval-aa GDPval-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. AA-3D: Interactive 3D Experience Benchmark https://artificialanalysis.ai/evaluations/aa-3d A benchmark developed by Artificial Analysis for how well AI models build interactive 3D experiences. Models work as agents in an offline sandbox via Stirrup to turn a written brief into a three.js scene, with Elo ratings derived from blind pairwise votes. APEX-Agents-AA Benchmark https://artificialanalysis.ai/evaluations/apex-agents-aa Artificial 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 https://artificialanalysis.ai/evaluations/aa-analyst-agent Artificial 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 Benchmark https://artificialanalysis.ai/evaluations/automationbench-aa A 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 v1.1 Benchmark https://artificialanalysis.ai/evaluations/harvey-lab-aa Artificial 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. v1.1, built in collaboration with Harvey, checks every deliverable against the task's source documents for hallucinations, so work that passes every rubric criterion but contains a material hallucination receives no credit. A panel of three judges grades each deliverable criterion-by-criterion. GDP.pdf Benchmark https://artificialanalysis.ai/evaluations/gdp-pdf Artificial 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 https://artificialanalysis.ai/evaluations/enterprise-ops-gym-aa Artificial 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 https://artificialanalysis.ai/evaluations/terminalbench-4-0 A 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 https://artificialanalysis.ai/evaluations/terminal-bench-science A 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. FrontierCode 1.1 Benchmark https://artificialanalysis.ai/evaluations/frontiercode Artificial Analysis' implementation of Cognition's FrontierCode 1.1, a private agentic coding benchmark that tests models on completing coding tasks in a way that would actually get merged by maintainers of real-world repositories. AA-PersonalityBench: AI Personality & Communication Style Benchmark https://artificialanalysis.ai/evaluations/aa-personalitybench Measures how models communicate rather than whether they are correct: 30 judged personality traits rolled up into 8 indexes, plus 21 measured communication characteristics. AA-LCR v1.1: Long Context Reasoning Benchmark https://artificialanalysis.ai/evaluations/artificial-analysis-long-context-reasoning A 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 Benchmark https://artificialanalysis.ai/evaluations/omniscience A benchmark measuring factual recall and hallucination across various economically relevant domains. SciCode Benchmark https://artificialanalysis.ai/evaluations/scicode A scientist-curated coding benchmark featuring 288 test set subproblems from 80 laboratory problems across 16 scientific disciplines. Humanity's Last Exam Benchmark https://artificialanalysis.ai/evaluations/humanitys-last-exam A frontier-level benchmark with 2,500 expert-vetted questions across mathematics, sciences, and humanities, designed to be the final closed-ended academic evaluation. GPQA Diamond Benchmark https://artificialanalysis.ai/evaluations/gpqa-diamond 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 https://artificialanalysis.ai/evaluations/itbench-aa Artificial 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 https://artificialanalysis.ai/evaluations/mmmu-pro An enhanced MMMU benchmark that eliminates shortcuts and guessing strategies to more rigorously test multimodal models across 30 academic disciplines. IFBench Benchmark https://artificialanalysis.ai/evaluations/ifbench A benchmark evaluating precise instruction-following generalization on 58 diverse, verifiable out-of-domain constraints that test models' ability to follow specific output requirements. MLCR-AA: Medical Long Context Reasoning Benchmark https://artificialanalysis.ai/evaluations/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. HealthBench Benchmark https://artificialanalysis.ai/evaluations/healthbench OpenAI's HealthBench: 5,000 health conversations graded against physician-written rubrics, from emergency referrals to global health. HealthBench Professional Benchmark https://artificialanalysis.ai/evaluations/healthbench-professional OpenAI's HealthBench Professional: 525 clinician conversations covering consult, writing and research tasks, graded against physician-written rubrics. CritPt v2 Benchmark https://artificialanalysis.ai/evaluations/critpt-v2 CritPt v2 is the revised version of CritPt, a benchmark designed to test LLMs on research-level physics reasoning tasks, featuring 70 composite research challenges. Models solve problems with a code-execution tool in an offline Linux sandbox. 𝜏³-Banking Benchmark https://artificialanalysis.ai/evaluations/tau3-banking A 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 https://artificialanalysis.ai/evaluations/terminalbench-2-1 A 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. CritPt Benchmark https://artificialanalysis.ai/evaluations/critpt A benchmark designed to test LLMs on research-level physics reasoning tasks, featuring 71 composite research challenges. Terminal-Bench Hard Benchmark https://artificialanalysis.ai/evaluations/terminalbench-hard An agentic benchmark evaluating AI capabilities in terminal environments through software engineering, system administration, and data processing tasks. 𝜏²-Bench Telecom Benchmark https://artificialanalysis.ai/evaluations/tau2-bench A 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 https://artificialanalysis.ai/evaluations/mmlu-pro An enhanced version of MMLU with 12,000 graduate-level questions across 14 subject areas, featuring ten answer options and deeper reasoning requirements. LiveCodeBench Benchmark https://artificialanalysis.ai/evaluations/livecodebench A 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 https://artificialanalysis.ai/evaluations/math-500 A 500-problem subset from the MATH dataset, featuring competition-level mathematics across six domains including algebra, geometry, and number theory. AIME 2025 Benchmark https://artificialanalysis.ai/evaluations/aime-2025 All 30 problems from the 2025 American Invitational Mathematics Examination, testing olympiad-level mathematical reasoning with integer answers from 000-999. Global-MMLU-Lite Benchmark https://artificialanalysis.ai/evaluations/global-mmlu-lite A lightweight, multilingual version of MMLU, designed to evaluate knowledge and reasoning skills across a diverse range of languages and cultural contexts.