Neura Intelligence Index
AI2 Reasoning Challenge — grade-school science questions requiring multi-step reasoning
| # | Model | Score | Provider | Source |
|---|---|---|---|---|
| 1 | 96.9 | Meta | ||
| 2 | 96.7 | Anthropic | ||
| 3 | 96.4 | Anthropic | ||
| 4 | 96.4 | OpenAI | ||
| 5 | 96.3 | OpenAI | ||
| 6 | 95.3 | Alibaba | ||
| 7 | 95.2 | Meta | ||
| 8 | 94.8 | Meta | ||
| 9 | 94.8 | NVIDIA | ||
| 10 | 94 | Mistral | ||
| 11 | 93.1 | OpenAI | ||
| 12 | 91.8 | Anthropic | ||
| 13 | 91.6 | Microsoft | ||
| 14 | 91.3 | Mistral | ||
| 15 | 88.4 | Cohere | ||
| 16 | 83.4 | Meta |