List your company

Audio Understanding on MMAU

81.1AccuracyGemini-3.1-Pro
0.08421.11742.1563.183Jan 26, 2026Feb 20, 2026Mar 18, 2026Apr 13, 2026May 8, 2026Jun 3, 2026Jun 29, 2026

Evaluation results

MethodLinks
Gemini-3.1-ProSize=–, Model Category=Closed-source models2026.0681.1
Gemini-3-Pro2026.0280.8
ERNIE 5.0-Base2026.0280.8
ERNIE 5.02026.0280.4
Gemini-3-ProSize=–, Model Category=Closed-source models2026.0680.15
Step-Audio-R1Size=33B, Model Category=Open-source models2026.0678.67
Qwen3-Omni-A3B-InstructVariant=Omni-A3B-Instruct2026.0378.2
MOSS-Audio-4B-ThinkingSize=4B, Model Category=Open-source models2026.0677.64
BaseBase Model=Qwen3-Omni-30B-A3B-Instruct2026.0577.6
Qwen3-Omni-Instruct2026.0277.5
MOSS-Audio-8B-ThinkingSize=8B, Model Category=Open-source models2026.0677.33
MOSS-Audio-8B-InstructSize=8B, Model Category=Open-source models2026.0677.03
ReGapBase Model=MiniCPM-o-4.52026.0576.8
ReGapBase Model=Qwen3-Omni-30B-A3B-Instruct2026.0576.8
LongCat-Next2026.0376.4
BaseBase Model=MiniCPM-o-4.52026.0576.2
Audio Flamingo NextSize=8B, Model Category=Open-source models2026.0676.1
LongCat-Flash-Omni2026.0275.9
MiMo-Audio2026.0375.8
MOSS-Audio-4B-InstructSize=4B, Model Category=Open-source models2026.0675.79
Qwen3-Omni-30B-A3B-InstructSize=30B, Model Category=Open-source models2026.0675
MiMo-Audio-7BSize=7B, Model Category=Open-source models2026.0674.9
Gemini-2.5-Flash-Lite-previewVersion=2.5, Tier=Flash-Lite-preview2026.0374.8
MiMo-Audio + SICL-AT3few-shot=true2026.0173.4
MiMo-Audio + SICL-AT2few-shot=true2026.0172.9
Qwen2.5-Omni-3BModel Scale=3B2026.0672.8
MiMo-Audio + Vanilla SICLfew-shot=true2026.0172.6
Kimi-AudioSize=7B, Model Category=Open-source models2026.0672.41
Step-Audio-R1.1Size=33B, Model Category=Open-source models2026.0672.18
Qwen2.5-Omni + SICL-AT3few-shot=true2026.0172.1
MiMo-Audio + SICL-AT1few-shot=true2026.0171.9
Gemini-3.1-Flash-Lite-previewVersion=3.1, Tier=Flash-Lite-preview2026.0371.7
Step-Audio-2-miniVersion=2-mini2026.0371.3
ShiftDCBase Model=Qwen3-Omni-30B-A3B-Instruct2026.0571.2
Qwen2.5-Omni + SICL-AT2few-shot=true2026.0171.1
MiniCPM-o-4.5Size=9B, Model Category=Open-source models2026.0670.97
Kimi-Audio2026.0370.31
Qwen2.5-Omni + SICL-AT1few-shot=true2026.0169.6
GPT-4o-Audio2026.0268.4
AVLM-2BModel Scale=2B2026.0668.3
Qwen2.5-Omni + Vanilla SICLfew-shot=true2026.0167.3
MiMo-Audiofew-shot=false2026.0166.9
ReGapBase Model=Qwen2.5-Omni-3B2026.0566.8
BaseBase Model=Qwen2.5-Omni-3B2026.0566
Qwen2.5-Omnifew-shot=false2026.0165.8
GPT4o-AudioSize=–, Model Category=Closed-source models2026.0665.66
Qwen2.5-OmniSize=7B, Model Category=Open-source models2026.0665.6
Qwen2.5-Omni (Fine-tuned on RSR)few-shot=false2026.0165.3
Kimi Audio2026.0265.2
SPA-VL-DPOBase Model=Qwen2.5-Omni-3B2026.0564.4
Qwen2.5-Omni (Fine-tuned on CV-en) + Vanilla SICLfew-shot=true2026.0164.1
VLGuard-SFTBase Model=Qwen2.5-Omni-3B2026.0564
Qwen2.5-Omni (Fine-tuned on CV-en)few-shot=false2026.0163.9
Audio Flamingo 3Size=7B, Model Category=Open-source models2026.0661.23
Gemma-4-E2B-itModel Scale=4B2026.0655.9
ShiftDCBase Model=MiniCPM-o-4.52026.0529.2
ShiftDCBase Model=Qwen2.5-Omni-3B2026.053.2