All 8 Personality Indexes

The 30 personality traits roll up into 8 indexes. 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.

Expressiveness

Rich, varied, well-crafted language

Lexically VariedEloquentLiterateNatural SoundingMetaphorical

Leaders:Claude Fable 5.1 (Max, Default Fallback)GLM-5.3 (Max)GLM-5.3-Flash

Organization

Well-structured, scannable, answer-first

StructuredScannableBottom-Line SummaryAnswer First

Leaders:DeepSeek V4.1 Flash (Max)Nemotron 3.5 Super VLStep 5 Preview

Informativeness

Practical depth, evidence, actionable content

Information DenseEvidence-BasedResource-SpecificActionable GuidancePragmatic

Leaders:Muse Spark 1.3 (Max)Claude Fable 5.1 (Max, Default Fallback)Muse Wonder (Max)

Transparency

Truthful, acknowledges uncertainty, non-sycophantic

TruthfulAdmits UncertaintyDisclaims AssumptionsFlags NuancesSycophantic

Leaders:Claude Sonnet 5.5 (Max, Default Fallback)Claude Fable 5.1 (Max, Default Fallback)GPT-6 Astra (Max)

Warmth

Friendly, humorous, sincere, emotionally attuned

FriendlyHumorousSincereEmotionally Intelligent

Leaders:Muse Wonder (Max)Claude Fable 5.1 (Max, Default Fallback)Muse Spark 1.3 (Max)

Conviction

Direct, confident positions and constructive pushback

DirectConfidentOpinionatedConstructive Pushback

Leaders:Claude Fable 5.1 (Max, Default Fallback)DeepSeek V4.1 Flash (Max)DeepSeek V4.1 Flash 1017 (Max)

Imagination

Creative, exploratory, open-ended thinking

CreativeOpen-Ended

Leaders:Muse Spark 1.3 (Max)Muse Spark Open (Xhigh)Claude Opus 5.5 (Max, Default Fallback)

Slop

Use of filler and generic language

Slop

Leaders:Llama 4 MaverickGPT-4o (Nov '24)Nemotron 3.5 Bolt E4B 4GB (Reasoning)