Humor Elo Leaderboard: Gemini Crushes Rivals While GPT-3.5 Turbo Tanks at 179
opmgyhx · reddit · 2026-10-20
A Reddit user shared Artificial Analysis's AA-PersonalityBench humor leaderboard, which measures whether models use humor, wit, or playful language when appropriate.
- Gemini 4 Argon leads by a huge margin at 1573 Elo, 157 points above second-place Gemini 3.8 Flash (1416), with Claude Fable 5.1, Muse Wonder, Gemini 3.1 Pro Preview, and Claude Opus 5.5 rounding out the top five
- Most mainstream models cluster between 700-1400, while GPT-3.5 Turbo sits dead last at 179, with Llama 4 Maverick and Nemotron 3 Ultra at just 490 and 532
- Scores come from pairwise verdicts by a panel of LLM judges across 30 prompts per trait
Personality-trait benchmarks like humor are emerging as a new ranking axis beyond reasoning and coding.
Related event: AA-PersonalityBench Humor Elo Ranking: Gemini 4.1 Arvon Tops the Chart(2 posts)→
More from Models
- Reddit user ditches US AI for Chinese models: cheaper, fewer refusals, no watermarks — Revolutionalredstone · 2026-10-21
- Fireworks, Microsoft and NVIDIA host SF buildathon: beat a frontier model, win $10K — sophiamyang · 2026-10-21
- OSWorld-Science benchmark: 12 VLMs average just 34.3% on real scientific software tasks — jiqizhixin · 2026-10-21
- Experiments show Qwen3 can compile arbitrary patterns into layer-by-layer steps with plan training — Sauers_ · 2026-10-21
- Hugging Face quietly added model file size display, a big QoL win — vboykis · 2026-10-21
- Humans& launches Ampersand, first model with privacy built in, plus a funny RL-plot joke — niloofar_mire · 2026-10-21