Audio-Language Model Evaluation on ALME
72nAUC (0-5 pp budget)GACL
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GACLModel=Voxtral-Small2026.06 | 72 | 76 | |
| GACLModel=Qwen2.5-Omni2026.06 | 66 | 66 | |
| GACLModel=Qwen3-Omni2026.06 | 64 | 64 | |
| GACLModel=Qwen2-Audio-Inst.2026.06 | 60.4 | 63.4 | |
| AADModel=Qwen2.5-Omni2026.06 | 54.8 | 56.4 | |
| GACLModel=Kimi-Audio-Inst.2026.06 | 47 | 47 | |
| AADModel=Qwen3-Omni2026.06 | 44.7 | 45 | |
| AADModel=Qwen2-Audio-Inst.2026.06 | 39.5 | 43.7 | |
| ACDModel=Qwen2.5-Omni2026.06 | 32.8 | 41.2 | |
| ACDModel=Qwen2-Audio-Inst.2026.06 | 21.5 | 26.8 | |
| AADModel=Voxtral-Small2026.06 | 17.9 | 29.4 | |
| ACDModel=Qwen3-Omni2026.06 | 5.2 | 5.2 | |
| ACDModel=Kimi-Audio-Inst.2026.06 | 2.2 | 4.1 | |
| ACDModel=Voxtral-Small2026.06 | 2.1 | 7 | |
| AADModel=Kimi-Audio-Inst.2026.06 | 2 | 4.2 |