LAB-Bench FigQA With Cropping (Opus 4.6 System Card) — leaderboard
Metric: Accuracy (%). Source: www-cdn.anthropic.com. 3 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 (Thinking) | 78.3 |
| 2 | Human Expert | 77 |
| 3 | Claude Opus 4.5 (Thinking) | 69.4 |
Interactive version: theaggregate.ai/benchmark?slug=lab-bench-figqa-with-cropping-opus-4-6-system-card · How the rankings work · Data refreshed daily, snapshot 2026-07-22.