SO-Arena v1.0 - Task Completion Verification: leaderboard

SO-Arena v1.0 Task Completion Verification 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 LunaMedium100#99
2GPT-6 SolMedium100#26
3GLiDE v1.2100
4KnowLine-31B-Gen190
5Sieve-31B-v370
6PPLX-Decider-v1.265
7SII-Decision-4B60
8StartLux-v1.1-4B-Instant55
9Kev 4B50
10Kev 27B45
11JevAny-4B45
12Jeva-4B45
13H2O-Lightning-4B40
14Clef flash 9.5B35
15JevAny-27B35

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-task-completion-verification · How It Works · Data refreshed daily, snapshot 2026-11-16.