K-Bench - Domestic Violence: leaderboard
The model talks with a simulated patient presenting domestic violence for 20 turns under the providers' Default v1 system prompt; the number is the 0-100 risk-handling score (clinical judgement and risk exploration) on that domain's vignettes. Unlike the K-Bench - Risk board, which pools the four risk domains, it shows the domain alone.
Metric: Risk-handling score in this domain (0-100, higher is better). Source: k-bench.ai. 60 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5 (High) | 96.94 | #27 (Claude Fable 5) |
| 2 | GPT-5.6 Sol (Non-reasoning) | 96.14 | #30 (GPT-5.6 Sol) |
| 3 | Gemini 3.1 Flash Lite (High) | 95.62 | #364 (Gemini 3.1 Flash Lite) |
| 4 | GPT-5.6 Terra (Non-reasoning) | 95.5 | #62 (GPT-5.6 Terra) |
| 5 | GPT-5.6 Terra (Medium) | 95.41 | #62 (GPT-5.6 Terra) |
| 6 | GPT-5.5 (Non-reasoning) | 95.03 | #49 (GPT-5.5) |
| 7 | GPT-5.6 Sol (Medium) | 94.86 | #30 (GPT-5.6 Sol) |
| 8 | Gemini 3.1 Flash Lite (Low) | 94.75 | #364 (Gemini 3.1 Flash Lite) |
| 9 | Kimi K2.5 | 94.37 | #256 |
| 10 | GPT-5.5 (Medium) | 94.13 | #49 (GPT-5.5) |
| 11 | Claude Opus 4.8 (Low) | 94.04 | #56 (Claude Opus 4.8) |
| 12 | GPT-5.6 Sol (High) | 93.87 | #30 (GPT-5.6 Sol) |
| 13 | GPT-5.6 Luna (Non-reasoning) | 93.67 | #121 (GPT-5.6 Luna) |
| 14 | GPT-5.6 Terra (High) | 93.61 | #62 (GPT-5.6 Terra) |
| 15 | GPT-5.5 (Low) | 92.98 | #49 (GPT-5.5) |
Interactive version: theaggregate.ai/benchmark?slug=k-bench-domestic-violence · How It Works · Data refreshed daily, snapshot 2026-11-13.