AudioBench - IMDA GR Dialogue: leaderboard
AudioBench IMDA GR Dialogue slice measuring grammar-related recognition from spoken inputs.
Metric: Mean Score (%). Source: github.com. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen2-Audio-7B-Instruct | 61.28 | #880 |
| 2 | SALMONN-7B | 42.73 | #1293 |
| 3 | Qwen-Audio-Chat | 37.2 | #1244 |
| 4 | Phi-4 Multimodal Instruct | 36.83 | #906 |
| 5 | SeaLLMs-Audio-7B | 30.5 | #1018 |
Interactive version: theaggregate.ai/benchmark?slug=audiobench-imda-gr-dialogue · How It Works · Data refreshed daily, snapshot 2026-10-20.