EduArt - Multiple Choice: leaderboard

Metric: Exact-match accuracy (%; the 370 single-answer multiple-choice items in Italian and English, answer-only condition, one run; higher is better). Source: arxiv.org. Saturation forecast: Estimated already saturated. 12 models tracked.

Top models

#ModelScore
1Gemini 3.5 Flash96.5
2Gemini 3.1 Pro (Preview)95.7
3GPT-5.595.1
4Claude Opus 4.694.6
5Claude Sonnet 4.694.3
6Gemini 3.1 Flash Lite (Preview)92.4
7Qwen 3 VL 235B A22B88.6
8Mistral Large 384.6
9Claude Haiku 4.584
10GPT-5.4 Mini82.7
11GPT-5.4 Nano62.7
12Pixtral Large28.6

Interactive version: theaggregate.ai/benchmark?slug=eduart-multiple-choice · How It Works · Data refreshed daily, snapshot 2026-09-29.