Artificial Analysis reports Kimi K3.1 Pro's personality and writing style differ from Kimi K3
Overview
Artificial Analysis reports that Kimi K3.1 Pro shows the largest personality change it has measured against a direct predecessor on AA-PersonalityBench.
Its percentile ranks moved an average of 30 points from Kimi K3's across 30 traits, compared with 23 points for GPT-6 Sol to GPT-6.1 Sol, and 18 traits shifted 25 points or more.
The firm says the largest changes are on Organization, which rises from 33 to 88, while Warmth falls from 93 to 51, Imagination from 88 to 53, Expressiveness from 93 to 63, and Informativeness from 51 to 23. It also says Kimi K3.1 Pro's default model selection clusters with GPT-6.1 Sol and GPT-6 Astra, while Kimi K3 clusters with GLM-5.3 and Claude Fable 5.1.
In writing style, Artificial Analysis says Kimi K3.1 Pro uses fewer parenthetical asides, em dashes, contractions, and italics than Kimi K3, and more bold text, headings, and tables.
Written by AI from the articles below · updated Nov 2, 5:41 PM ET
Check the sources:
Developments
4 developments
- Nov 2, 5:38 PM ET · 1 articleArtificial Analysis reports Kimi K3.1 Pro's writing style differs from Kimi K3Artificial Analysis: Kimi K3.1 Pro shifts writing style toward bold text, headings, and tables
- Nov 2, 5:38 PM ET · 1 articleArtificial Analysis compares Kimi K3.1 Pro with Kimi K3 on its writing-quality indexArtificial Analysis: Kimi K3.1 Pro gains on Organization but trails K3 on Warmth and Imagination
- Nov 2, 5:38 PM ET · 1 articleArtificial Analysis adds Kimi K3 to default model selection, clusters K3.1 Pro with GPT modelsArtificial Analysis: Kimi K3.1 Pro clusters with GPT-6 models, unlike Kimi K3
- Nov 2, 5:38 PM ET · 1 articleArtificial Analysis reports Kimi K3.1 Pro shows largest personality change vs. K3 on AA-PersonalityBenchArtificial Analysis: Kimi K3.1 Pro shows largest personality shift from Kimi K3
Article timeline
The articles in this story. Times are ET.
Artificial Analysis@ArtificialAnlysOfficialKimi K3.1 Pro shifts writing style toward bold text, headings, and tablesAIArtificial Analysis says Kimi K3.1 Pro's responses use fewer parenthetical asides, em dashes, contractions, and italics than Kimi K3. The model uses more bold text, headings, and tables instead.

Artificial Analysis@ArtificialAnlysOfficialKimi K3.1 Pro gains on Organization but trails K3 on Warmth and ImaginationAIArtificial Analysis says Kimi K3.1 Pro ranks higher than Kimi K3 on Organization, rising from 33 to 88. It ranks lower on Warmth (93 to 51), Imagination (88 to 53), Expressiveness (93 to 63), and Informativeness (51 to 23).

Artificial Analysis@ArtificialAnlysOfficialKimi K3.1 Pro clusters with GPT-6 models, unlike Kimi K3AIArtificial Analysis says Kimi K3.1 Pro falls in a different personality family from Kimi K3 in its default model selection. Kimi K3 is most similar to GLM-5.3 and Claude Fable 5.1, while Kimi K3.1 Pro is most similar to GPT-6.1 Sol and GPT-6 Astra.

Artificial Analysis@ArtificialAnlysOfficialKimi K3.1 Pro shows largest personality shift from Kimi K3AIArtificial Analysis reports that Kimi K3.1 Pro shows the largest personality change it has measured against a direct predecessor on AA-PersonalityBench. Across 30 traits, its percentile ranks moved by an average of 30 points from Kimi K3's, versus 23 for GPT-6 Sol to GPT-6.1 Sol, and 18 traits shifted 25 points or more.

Heat Trend
Not enough continuous observations to show a trend yet.