Creative Plot Generation on 160 prompts NQD (test)
8.67Character DevelopmentClaude Sonnet 4
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Claude Sonnet 42026.03 | 8.67 | 8.64 | 8.65 | 8.81 | 8.95 | |
| PLOTTWIST2026.03 | 8.64 | 8.7 | 8.85 | 8.89 | 8.98 | |
| Agents’ Room2026.03 | 8.55 | 8.59 | 8.68 | 8.81 | 8.94 | |
| Gemini 2.0 Flash2026.03 | 8.39 | 8.55 | 8.67 | 8.72 | 8.89 | |
| GPT-4.12026.03 | 8.29 | 8.61 | 8.74 | 8.74 | 8.88 | |
| Qwen3-32B2026.03 | 8.25 | 8.53 | 8.5 | 8.36 | 8.87 | |
| Mistral Small 2501 24B2026.03 | 8.07 | 8.36 | 8.48 | 8.38 | 8.77 | |
| Qwen2.5-14B-Instruct2026.03 | 8.03 | 8.37 | 8.48 | 8.47 | 8.8 | |
| DeepSeek R1 14B2026.03 | 7.92 | 8.36 | 8.44 | 8.4 | 8.76 | |
| Qwen2.5-7B-Instruct2026.03 | 7.88 | 8.37 | 8.37 | 8.22 | 8.7 | |
| Llama 70B2026.03 | 7.84 | 8.27 | 8.22 | 8.11 | 8.68 | |
| Phi-4 Mini Instruct2026.03 | 7.45 | 7.97 | 7.81 | 7.51 | 8.35 | |
| WizardLM-StoryTelling-30B2026.03 | 7.41 | 8.14 | 8.2 | 7.89 | 8.38 |