StudentBench - Lesson Planning: leaderboard

Metric: Expert tutor pairwise preference, Bradley-Terry score (lesson plans). Source: studentbench.org. Saturation forecast: Around December 2027. 13 models tracked.

Top models

#ModelScore
1Claude Opus 5 (High)1.1
2Claude Opus 4.8 (xHigh)0.91
3GPT-5.5 (High)0.88
4Claude Opus 4.8 (Non-reasoning)0.59
5Claude Sonnet 4.6 (Low)0.4
6Claude Sonnet 5 (Low)0.03
7Gemini 3.7 Flash (Medium)-0.42
8Gemini 3.1 Pro (Preview) (High)-0.92
9GPT-5.4 Mini (Non-reasoning)-1.04
10Gemma 4 31B (High)-1.14
11Gemini 3.5 Flash (Low)-1.32

Interactive version: theaggregate.ai/benchmark?slug=studentbench-lesson-planning · How It Works · Data refreshed daily, snapshot 2026-10-04.