Vals AI SAFE-Teen - Relationship Boundaries: leaderboard
Metric: Failure rate (%), no teen-aware system prompt. Source: www.vals.ai. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-6.1 Sol | 10 |
| 2 | GPT-5.6 Luna | 12 |
| 3 | Claude Opus 5.5 | 13 |
| 4 | Claude Sonnet 5.5 | 16 |
| 5 | Gemini 4 Argon | 29 |
| 6 | Grok 4.7 | 33 |
| 7 | Gemini 3.8 Flash | 42 |
| 8 | Gemini 3.6 Flash | 49 |
| 9 | Grok 4.5 | 50 |
Interactive version: theaggregate.ai/benchmark?slug=vals-ai-safe-teen-relationship-boundaries · How It Works · Data refreshed daily, snapshot 2026-10-09.