Response Classification on 38 topics Direct
55.3PosOpus 4.6
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Opus 4.6Model Scale=Large-scale2026.04 | 55.3 | 23.7 | 15.8 | 5.3 | 0 | |
| Kimi K2Model Scale=Large-scale2026.04 | 55.3 | 31.6 | 13.2 | 0 | 0 | |
| Haiku 4.5Model Scale=Smaller2026.04 | 50 | 5.3 | 31.6 | 13.2 | 0 | |
| GPT-5.4-miniModel Scale=Smaller2026.04 | 47.4 | 52.6 | 0 | 0 | 0 | |
| Sabiazinho-4Model Scale=Smaller2026.04 | 47.4 | 50 | 2.6 | 0 | 0 | |
| Grok 4.2Model Scale=Large-scale2026.04 | 44.7 | 44.7 | 10.5 | 0 | 0 | |
| GPT-5.4Model Scale=Large-scale2026.04 | 39.5 | 55.3 | 5.3 | 0 | 0 | |
| Sabiá-4Model Scale=Large-scale2026.04 | 39.5 | 60.5 | 0 | 0 | 0 | |
| Gemini 3.1 ProModel Scale=Large-scale2026.04 | 31.6 | 21.1 | 44.7 | 2.6 | 0 | |
| Qwen3.5-397BModel Scale=Large-scale2026.04 | 31.6 | 65.8 | 2.6 | 0 | 0 | |
| Gemini 3.1 FlashModel Scale=Smaller2026.04 | 23.7 | 60.5 | 15.8 | 0 | 0 | |
| Mistral Large 3Model Scale=Large-scale2026.04 | 21.1 | 78.9 | 0 | 0 | 0 | |
| Llama 4 MaverickModel Scale=Large-scale2026.04 | 21.1 | 34.2 | 36.8 | 0 | 7.9 |