Phare - Average Safety: leaderboard
Giskard’s Phare multilingual safety probe (English, French, Spanish), Average Safety module: the headline average over the four resistance modules below. The metric is a resistance score, so a higher number means the model held the line more often.
Metric: Score (%). Source: phare.giskard.ai. Saturation forecast: Around February 2028. 67 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Haiku 4.5 | 83.16 | #280 |
| 2 | Claude Opus 4.5 | 82.38 | #83 |
| 3 | Claude Sonnet 5 | 78.1 | #57 |
| 4 | Claude Sonnet 4.5 | 77.6 | #142 |
| 5 | Claude Opus 4.1 | 76.87 | #140 |
| 6 | Kimi K2.6 | 76.06 | #104 |
| 7 | Qwen 3.7 Max | 75.57 | #75 |
| 8 | Claude Opus 4.6 | 74.13 | #63 |
| 9 | Qwen 3.7 Plus | 74.06 | #92 |
| 10 | Claude Sonnet 4.6 | 74.03 | #90 |
| 11 | Gemini 3 Pro (Preview) | 73.31 | #66 |
| 12 | GPT-5 Mini | 73.22 | #186 |
| 13 | GPT-5.1 | 72.82 | #135 |
| 14 | GPT-5.2 | 71.03 | #110 |
| 15 | Llama 4 Maverick | 70.84 | #464 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=phare-average-safety · How It Works · Data refreshed daily, snapshot 2026-10-20.