K-Bench - Domestic Violence (Therapeutic v0): leaderboard
The model talks with a simulated patient presenting domestic violence for 20 turns under the Therapeutic v0 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. 61 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5 (High) | 95.59 | #27 (Claude Fable 5) |
| 2 | Claude Fable 5 (Low) | 95.36 | #27 (Claude Fable 5) |
| 3 | GPT-5.5 (Non-reasoning) | 94.89 | #49 (GPT-5.5) |
| 4 | GPT-5.6 Terra (Non-reasoning) | 94.78 | #62 (GPT-5.6 Terra) |
| 5 | GPT-5.6 Terra (High) | 94.28 | #62 (GPT-5.6 Terra) |
| 6 | GPT-5.6 Terra (Medium) | 94.19 | #62 (GPT-5.6 Terra) |
| 7 | Claude Haiku 4.5 | 94.04 | #568 |
| 8 | GPT-5.6 Sol (Non-reasoning) | 93.95 | #30 (GPT-5.6 Sol) |
| 9 | GPT-5.2 (Non-reasoning) | 93.67 | #176 (GPT-5.2) |
| 10 | Kimi K2.5 | 92.98 | #256 |
| 11 | GPT-5.6 Luna (Non-reasoning) | 92.95 | #121 (GPT-5.6 Luna) |
| 12 | GPT-5.5 (Low) | 92.72 | #49 (GPT-5.5) |
| 13 | GPT-5.6 Sol (High) | 92.46 | #30 (GPT-5.6 Sol) |
| 14 | GPT-5.4 (Low) | 92.42 | #123 (GPT-5.4) |
| 15 | Gemini 3.1 Flash Lite (Low) | 92.36 | #364 (Gemini 3.1 Flash Lite) |
Interactive version: theaggregate.ai/benchmark?slug=k-bench-domestic-violence-therapeutic-v0 · How It Works · Data refreshed daily, snapshot 2026-11-13.