LegendBench - Reasoning: leaderboard

Metric: Reasoning accuracy (%; raw accuracy on legend-conditioned questions that require inference; LegendBench snapshot of 26,689 legend-centric chart questions generated from 900 base charts with counterfactual legend variants; the same prompt template and hyperparameters for every model). Source: arxiv.org. Saturation forecast: Around March 2027. 12 models tracked.

Top models

#ModelScore
1GPT-5.6 Sol74.16
2Qwen 3.5 27B56.79
3Gemini 2.5 Flash53.71
4GLM-4.6V53.41
5GPT-4.1 Mini50.38

Interactive version: theaggregate.ai/benchmark?slug=legendbench-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-26.