AI for Education SEND: leaderboard
A 223-item set of four-option multiple choice questions on teaching pupils with special educational needs and disabilities, drawn from the special education papers of Chile's ECEP teacher exams. Models answer under a fixed three-shot prompt whose example questions are left out of scoring, and the chosen letter is marked by exact match. The number is accuracy, higher is better. Source: arxiv.org/abs/2506.18710.
Metric: Accuracy (%). Source: www.fab-ai.org. Saturation forecast: Around January 2027. 241 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5.1 | 89.45 | #9 |
| 2 | Claude Opus 5 | 88.99 | #12 |
| 3 | Gemini 3.6 Flash | 88.53 | #51 |
| 4 | GPT-5.5 | 88.07 | #31 |
| 5 | Qwen 3.8 Max | 88.07 | #35 |
| 6 | Muse Spark 1.2 | 88.07 | #62 |
| 7 | GPT-6 Astra (High) | 88.07 | #6 (GPT-6 Astra) |
| 8 | GPT-5.6 Sol | 87.61 | #19 |
| 9 | Claude Fable 5 | 87.16 | #16 |
| 10 | Gemini 3.7 Flash | 87.16 | #23 |
| 11 | Gemini 3.1 Pro (Preview) | 86.7 | #58 |
| 12 | Gemini 3.5 Flash | 86.24 | #60 |
| 13 | Gemini 3.8 Flash | 86.24 | #21 |
| 14 | Grok 4.6 | 86.24 | #26 |
| 15 | Qwen3.8 2.4T A95B | 86.24 | #55 |
No result here: #3 Claude Opus 5.5, #7 GPT-6.1 Sol, #8 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=ai-for-education-send · How It Works · Data refreshed daily, snapshot 2026-10-20.