K12-Bench: leaderboard
Large language models (LLMs) are increasingly used in K-12 education, yet existing benchmarks such as C-Eval, CMMLU, GaokaoBench, and EduEval mainly evaluate factual recall through exam-style question answering. Effective educational AI additionally requires curriculum cognition: understanding how knowledge is structured through prerequisite chains, concept taxonomies, experiment-concept links, an
Metric: Overall EM (self-reported). Source: benchmarklist.com. Saturation forecast: Around June 2027. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash | 57.1 | #99 |
| 2 | Gemini 2.5 Flash | 48.3 | #245 |
| 3 | Gemma 4 31B (IT) | 46.4 | #176 |
| 4 | GPT-5.2 | 42.8 | #110 |
| 5 | Qwen 3 32B | 42.6 | #443 |
| 6 | GPT-5 Mini | 31.7 | #186 |
| 7 | GLM-4.7 Flash | 31.7 | #509 |
| 8 | GPT-4o | 31.1 | #343 |
| 9 | Llama 3 8B Instruct | 7.2 | #1129 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=k12-bench · How It Works · Data refreshed daily, snapshot 2026-10-20.