EQ-Bench 3: Insight
Chat & Writing · 2025-04-28
Within EQ-Bench 3's emotional-intelligence scenarios, this subscore rates the insight dimension of model responses on a 0–10 scale. Higher values are better.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Fable 5 | 9.4 |
| Kimi K2.6 | 9.3 |
| GPT-5.4 | 9.2 |
| Opus 4.7 | 9.1 |
| Opus 4.8 | 9.1 |
| GPT-5.5 | 9.0 |
| DeepSeek-V4-Pro | 8.9 |
| Opus 4.6 | 8.8 |
| Sonnet 4.5 | 8.8 |
| GLM-5.1 | 8.8 |
| GLM-5.2 | 8.8 |
| GPT-5.2 | 8.8 |
| Sonnet 4.6 | 8.7 |
| Kimi K2.5 | 8.7 |
| Kimi K2 Instruct | 8.7 |
| Hivemind-32B-Preview | 8.6 |
| GLM-5 | 8.5 |
| Opus 4.5 | 8.4 |
| Gemini 3.1 Pro Preview | 8.4 |
| Gemini 3 Pro Preview | 8.4 |
| GPT-5.1 | 8.4 |
| Horizon Alpha | 8.4 |
| o3 | 8.3 |
| DeepSeek-V4-Flash | 8.1 |
| GPT-5 Chat (2025-08-07) | 8.1 |
| Qwen3.5 397B A17B | 8.1 |
| Gemini 2.5 Pro Preview 06-05 | 8.0 |
| GPT-5.3 Instant | 8.0 |
| Opus 4 | 7.9 |
| Sonnet 4 | 7.9 |
| GLM-4.7 | 7.9 |
| Gemma 4 26B A4B IT | 7.8 |
| Gemma 4 31B IT | 7.7 |
| DeepSeek-R1 | 7.6 |
| GLM-4.5 | 7.6 |
| Grok 4.1 Fast | 7.5 |
| Gemini 2.5 Pro Preview 03-25 | 7.3 |
| GPT-4.1 | 7.2 |
| DeepSeek-V3-0324 | 7.1 |
| ChatGPT-4o (2025-04-25 update) | 7.1 |