List your company

Automated Audio Captioning on MACS

64.9AAC ScoreKimi-Audio
1.51217.968534.42550.8815Sep 24, 2025

Evaluation results

MethodLinks
Kimi-AudioSize=7B2025.0964.9
SLAM-OmniSize=0.5B2025.0950.46
TtTSize=3B2025.0948.87
Qwen2.5-3B (AR)Size=3B2025.0948.64
Pretrain+TtTSize=3B2025.0942.86
Qwen2.5-3B (NAR)Size=3B2025.0927.4
GLM-4-VoiceSize=9B2025.0912.67
MoshiSize=7B2025.0912.01
VITA-AudioSize=7B2025.097.94
LLaMA-OmniSize=8B2025.094.56
Mini-OmniSize=0.5B2025.094.45
SpeechGPTSize=7B2025.093.95