HalluAudio - Music: leaderboard
Hallucination detection in the music domain of HalluAudio (2026): human-verified audio question-answer pairs with adversarial prompts, mixed audio and unanswerable questions probe whether an audio-language model reports sounds or content that are not there. Unweighted mean accuracy over nine task types, in percent; higher is better.
Metric: Average classification accuracy (%) over the nine music task types of HalluAudio (human-verified audio question-answer pairs with adversarial, mixed-audio and unanswerable prompts; zero-shot; outputs normalized to labels, refusals counted wrong); unweighted mean of the task accuracies; higher is better. Source: arxiv.org. Saturation forecast: Around 2031. 10 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | MiMo-Audio-7B-Instruct | 75.79 | #1328 |
| 2 | GPT-4o Audio (Preview) | 58.15 | #896 |
| 3 | Audio Flamingo 3 | 58.1 | |
| 4 | Step-Audio-2-mini | 58.1 | |
| 5 | Music Flamingo | 56.3 | #865 |
| 6 | Qwen2.5-Omni-7B | 50.96 | #1888 |
| 7 | Kimi-Audio-7B | 45.38 | |
| 8 | Gemini 2.5 Flash | 31.51 | #466 |
| 9 | Pengi | 28.38 | |
| 10 | Qwen2-Audio-7B | 27.56 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=halluaudio-music · How It Works · Data refreshed daily, snapshot 2026-11-22.