ChartBench - Value Comparison: leaderboard
ChartBench Value Comparison (Xu et al., 2023): zero-shot yes/no questions on whether a stated comparison between plotted values holds. Scored with Acc+, the percentage of questions where the model accepts the true assertion and rejects its paired false one, so guessing scores near zero; higher is better.
Metric: Acc+ (%). Source: chartbench.github.io. Saturation forecast: Around February 2027. 21 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4o | 66.19 | #741 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=chartbench-value-comparison · How It Works · Data refreshed daily, snapshot 2026-11-12.