EQ-Bench 3 (Rubric Score)
Chat & Writing · 2025-04-28
EQ-Bench 3 aggregate rubric score, a separate absolute 0-100 score from judge-scored scenario criteria.
Top models (higher is better)
| Model | Score |
|---|---|
| Opus 4.8 | 83.3 |
| Claude Fable 5 | 82.6 |
| Opus 4.7 | 82.6 |
| GPT-5.4 | 82.4 |
| GPT-5.5 | 82.4 |
| GPT-5.2 | 80.4 |
| Sonnet 4.6 | 80.0 |
| Hivemind-32B-Preview | 79.6 |
| Opus 4.6 | 79.2 |
| GLM-5.2 | 78.7 |
| GPT-5.1 | 78.1 |
| DeepSeek-V4-Pro | 77.6 |
| Opus 4.5 | 76.8 |
| Horizon Alpha | 76.8 |
| Kimi K2.6 | 76.6 |
| Gemini 3 Pro Preview | 75.8 |
| GLM-5.1 | 75.2 |
| GLM-5 | 75.0 |
| Kimi K2 Instruct | 75.0 |
| o3 | 74.7 |
| Gemini 3.1 Pro Preview | 74.3 |
| GPT-5 Chat (2025-08-07) | 73.7 |
| DeepSeek-V4-Flash | 72.5 |
| Kimi K2.5 | 72.5 |
| Sonnet 4.5 | 72.2 |
| Gemini 2.5 Pro Preview 06-05 | 72.2 |
| Qwen3.5 397B A17B | 71.3 |
| Gemma 4 31B IT | 70.8 |
| GPT-5.3 Instant | 70.7 |
| Gemma 4 26B A4B IT | 70.0 |
| GLM-4.7 | 68.9 |
| Opus 4 | 68.2 |
| ChatGPT-4o Latest observed 2025-03-27 | 67.8 |
| GLM-4.5 | 67.4 |
| Gemini 2.5 Pro Preview 03-25 | 67.3 |
| ChatGPT-4o (2025-04-25 update) | 66.7 |
| Sonnet 4 | 66.6 |
| DeepSeek-R1 | 65.8 |
| Hermes 4 405B | 65.2 |
| Grok 4.1 Fast | 65.1 |