SO-Arena v1.0 - Sentiment & Emotion Understanding: leaderboard

SO-Arena v1.0 Sentiment & Emotion Understanding domain score on the source-selected1232-task panel. Domain scores average native task credits for static decisions, tool outcomes or web progress, then scale to0–100. The full release has2919tasks; this leaderboard emphasizes the highest difficulty tiers and keeps all untiered sentiment/knowledge tasks. Metric: Native panel task credit mean (%). Higher is better.

Metric: Native panel task credit mean (%). Source: huggingface.co. Saturation forecast: Around February 2027. 20 models tracked. Updated 14 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1GLiDE v1.289.5
2PPLX-Decider-v1.288.75
3KnowLine-31B-Gen188.25
4PPLX-Decider-Flash-v1.288
5Sieve-31B-v388
6GPT-6 SolMedium87.75#26
7StartLux-v1.1-4B-Instant87.75
8Jev 1.14.085.5
9H2O-Lightning-4B85.25
10Clef 27B85
11GPT-6 LunaMedium84.75#99
12Kev 27B84.25
13JevAny-27B84.25
14StartLux-v1.1-4B-AutoThink83.75
15JevAny-4B83.25

No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.

Interactive version: theaggregate.ai/benchmark?slug=so-arena-v1-0-sentiment-emotion-understanding · How It Works · Data refreshed daily, snapshot 2026-11-16.