SO-Arena v1.0 - Intent & Routing: leaderboard

SO-Arena v1.0 Intent & Routing 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
1GPT-6 SolMedium82.29#26
2GPT-6 LunaMedium81.48#99
3KnowLine-31B-Gen178.19
4GLiDE v1.275.67
5StartLux-v1.1-4B-Instant70.2
6StartLux-v1.1-4B-AutoThink69.85
7PPLX-Decider-v1.269.84
8Kev 27B68.13
9Sieve-31B-v368.02
10Clef flash 9.5B64.33
11Clef 27B63.92
12H2O-Lightning-4B63.05
13Jeva-4B61.57
14JevAny-4B60.87
15PPLX-Decider-Flash-v1.259.39

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-intent-routing · How It Works · Data refreshed daily, snapshot 2026-11-16.