AudioBench - Public SG Speech QA: leaderboard
AudioBench Public SG Speech QA slice measuring question answering over spoken or audio content.
Metric: Mean Score (%). Source: github.com. Saturation forecast: Estimated already saturated. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Phi-4 Multimodal Instruct | 68.4 | #906 |
| 2 | Qwen-Audio-Chat | 63.17 | #1244 |
| 3 | Qwen2-Audio-7B-Instruct | 60.52 | #880 |
| 4 | SALMONN-7B | 59.24 | #1293 |
| 5 | SeaLLMs-Audio-7B | 57.59 | #1018 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=audiobench-public-sg-speech-qa · How It Works · Data refreshed daily, snapshot 2026-10-20.