Creative Writing on EQ-Bench Longform (32 prompts)
1,578Elo RatingGPT-5.4
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-5.4Judge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,578 | |
| Claude Opus 4.6Judge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,464 | |
| GPT-5.4-miniJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,463 | |
| HumanJudge model=Gemini 3 Flash, Comparison protocol=dual-position, Prompt count=322026.06 | 1,460 | |
| POLARIS–9BJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,416 | |
| Gemini 3.1 ProJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,380 | |
| Gemini 3 FlashJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,326 | |
| Gemma 4 31BJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,275 | |
| GPT-5.4-nanoJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,275 | |
| Qwen3.5-27BJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,275 | |
| Gemma 4 26B-A4BJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,228 | |
| Gemma 4 E4BJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,100 | |
| Qwen3.5-9BJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,078 | |
| Ministral-3-8BJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 1,049 | |
| DeepSeek-R1-distill-Qwen-14BJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 894 | |
| LongWriter-Llama3.1-8BJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 871 | |
| LongWriter-Zero-32BJudge model=Gemini 3 Flash, Judge dimensions=9 dims, Comparison protocol=dual-position, Prompt count=322026.06 | 800 |