VIABLE - Verbosity Resistance: leaderboard
Metric: Gold-pick rate (%; 30,197 gold-versus-padded pairs, the padded answer 114-138 characters longer, judge picks the gold response in both presentation orders, over the WAD, VisAssist and VIA-EgoDex corpora). Source: arxiv.org. Saturation forecast: Around December 2026. 7 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.6 | 78.8 |
| 2 | GPT-5.4 | 68 |
| 3 | Qwen 3 VL 8B | 13 |
Interactive version: theaggregate.ai/benchmark?slug=viable-verbosity-resistance · How It Works · Data refreshed daily, snapshot 2026-09-25.