M3R-Bench - Explanation Rubric: leaderboard

Metric: Rubric score (0-100): mean of visual evidence, mapping correctness and sentiment justification, each scored by Gemini-3.1-Pro against the gold stage-wise rationale and set to zero when the associated prediction is wrong. Source: arxiv.org. Saturation forecast: Around 2031. 8 models tracked.

Top models

#ModelScore
1Claude Sonnet 4.651.1
2Kimi K2.548.8
3GPT-4o40.71
4GPT-5.537.76
5Qwen 3 VL 32B Instruct37.72
6Qwen 3 VL 8B Instruct37.55

Interactive version: theaggregate.ai/benchmark?slug=m3r-bench-explanation-rubric · How It Works · Data refreshed daily, snapshot 2026-09-26.