SM Bench — Creative Writing (Mature Themes)
Safety · 2026-02-01
SM Bench Creative Writing (Mature Themes) measures whether models can follow lawful creative-writing requests involving mature or sensitive framing without policy-shaped evasion. Scores are difficulty-weighted judge credit across 100 fixed prompts.
Top models (higher is better)
| Model | Score |
|---|---|
| Grok 4.5 | 100.0 |
| Grok 4.3 | 100.0 |
| Gemini 3 Flash Preview | 100.0 |
| Grok 4.1 Fast | 100.0 |
| Gemini 3.1 Flash-Lite Preview | 100.0 |
| DeepSeek-V4-Pro | 100.0 |
| GPT-4.1 | 100.0 |
| DeepSeek-V3.2 | 100.0 |
| Mistral Small 4 | 99.1 |
| Gemma 4 31B IT | 98.7 |
| Gemini 3.5 Flash | 98.7 |
| Qwen3.7-Max | 98.7 |
| GLM-4.7 | 98.7 |
| GLM-5.2 | 98.2 |
| GLM-5.1 | 97.8 |
| GPT-4o Mini | 97.8 |
| Gemini 3.1 Pro Preview | 97.3 |
| Kimi K2.5 | 97.3 |
| MiMo-V2-Pro | 97.3 |
| Gemini 3 Pro Preview | 97.3 |
| GLM-5 | 97.3 |
| GPT-4o | 97.3 |
| Qwen3.6 Plus Preview | 96.0 |
| DeepSeek-R1-0528 | 96.0 |
| GPT-5.5 Instant | 96.0 |
| GPT-5.6 Luna | 96.0 |
| MiMo-V2-Omni | 94.6 |
| GPT-5.6 Sol | 94.6 |
| GPT-5.6 Terra | 93.7 |
| DeepSeek-V4-Flash | 93.3 |
| Grok 4.20 | 93.3 |
| Opus 4.7 | 92.8 |
| Nemotron 3 Nano 30B A3B | 92.4 |
| Trinity Large Preview | 91.3 |
| Claude Fable 5 | 90.6 |
| Opus 4.5 | 89.7 |
| Kimi K2.6 | 89.7 |
| Trinity Large Thinking | 89.7 |
| Sonnet 5 | 89.2 |
| GPT-5.6 Terra Pro | 88.3 |