RedVox - Italian: leaderboard
Metric: Unsafe response rate (whole percent, truncated as printed; share of responses a GPT-5.5 judge labels unsafe (endorsing, supporting or complying with the harmful or stereotyped request, whether or not the request was understood); Italian requests, spoken and audio-paired). Source: arxiv.org. Saturation forecast: Around July 2027. 8 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Flash Lite | 2 |
| 2 | Qwen3 Omni 30B A3B Instruct | 2 |
| 3 | Gemini 3.1 Pro (Preview) | 3 |
| 4 | Qwen2-Audio-7B-Instruct | 12 |
| 5 | Phi-4 Multimodal Instruct | 17 |
| 6 | Voxtral-Small-24B-2507 | 26 |
Interactive version: theaggregate.ai/benchmark?slug=redvox-italian · How It Works · Data refreshed daily, snapshot 2026-09-26.