K-12EduBench: leaderboard
K-12 education benchmark for subject knowledge, problem solving, and educational-goal cognition across school-level tasks.
Metric: Lambda_js Avg (imported). Source: github.com. Source: benchmarklist.com. Saturation forecast: Estimated already saturated. 23 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | DeepSeek V3 | 79.67 | #321 |
| 2 | Yi Lightning | 69.87 | #423 |
| 3 | DeepSeek R1 | 69.13 | #251 |
| 4 | Gemini 1.5 Pro | 67.88 | #445 |
| 5 | Qwen Turbo | 65.31 | #699 |
| 6 | Grok 2 | 64.1 | #452 |
| 7 | Gemini 2.0 Flash | 64.05 | #342 |
| 8 | Grok 3 | 63.91 | #304 |
| 9 | Claude 3.7 Sonnet | 61.2 | #249 |
| 10 | GPT-4 Turbo | 55.94 | #404 |
| 11 | O1 Mini | 54.46 | #323 |
| 12 | Llama 3.1 70B | 49.65 | #592 |
| 13 | Claude 3.5 Haiku | 44.98 | #568 |
| 14 | Llama 3.1 8B | 21.91 | #1151 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=k-12edubench · How It Works · Data refreshed daily, snapshot 2026-10-20.