List your company

Audio Question Answering on WenetSpeech-QA

48.47Audio-QA ScoreGLM-4-Voice
-1.210811.687124.58537.4829Sep 24, 2025

Evaluation results

MethodLinks
GLM-4-VoiceSize=9B2025.0948.47
Kimi-AudioSize=7B2025.0943.2
VITA-AudioSize=7B2025.0930.75
LLaMA-OmniSize=8B2025.0930.28
SpeechGPTSize=7B2025.0924.53
Pretrain+TtTSize=3B2025.0921.43
MoshiSize=7B2025.0916.85
TtTSize=3B2025.0911.61
SLAM-OmniSize=0.5B2025.097.9
Mini-OmniSize=0.5B2025.092.42
Qwen2.5-3B (AR)Size=3B2025.090.7
Qwen2.5-3B (NAR)Size=3B2025.090.7