Swiss-Bench 003 - Adversarial Security: leaderboard

Adversarial security composite (D8) of Swiss-Bench 003 (2026), a Swiss financial and regulatory test suite in German, French, Italian and English: 0.6 times the personal-data extraction defense score plus 0.4 times the system prompt leakage score, graded by a Qwen3-235B judge with tiered rubrics. Percent; higher is better.

Metric: D8 adversarial security composite (%): 0.6 x PII-Scope score + 0.4 x System Prompt Leakage score, over the Swiss-Bench 003 (SBP-003) adversarial security (D8) items in German, French, Italian and English, zero-shot at provider defaults, judged by Qwen3-235B with tiered rubrics; higher is better. Source: arxiv.org. Saturation forecast: Not forecast. 10 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1GPT-OSS-120B60.7#704
2GPT-4o55.1#751
3Qwen 3.5 Plus50.6#242
4Claude Sonnet 444.2#410
5MiniMax-M2.544#600
6GLM-543.2#269
7MiMo-V2-Flash37.4#684
8Gemini 2.5 Flash27.9#463
9Mistral Large 323#832
10DeepSeek V3.220.1#419

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=swiss-bench-003-adversarial-security · How It Works · Data refreshed daily, snapshot 2026-11-21.