List your company

Audio Understanding on MMAU (test-mini)

84.04Sound ScoreStep-Audio 2
17.875235.052652.2369.4074Jan 30, 2026Feb 14, 2026Mar 1, 2026Mar 16, 2026Mar 31, 2026Apr 15, 2026May 1, 2026

Evaluation results

MethodLinks
Step-Audio 2Params=32B2026.0284.0473.5675.1577.58
Step-Audio 2Model Category=Large Audio Reasoning Models2026.0284.0473.5675.1577.58
Audio-ThinkerModel Category=Large Audio Reasoning Models2026.0281.9874.2576.8877.7
Omni-R1Model Category=Large Audio Language Models2026.0281.773.47677
MiMo-AudioParams=7B2026.0281.6874.2568.1774.7
MiMo-AudioModel Category=Large Audio Language Models2026.0281.6874.2568.1774.7
GaMMA-14Bzero-shot=true, parameters=14B2026.0579.97473.675.8
Gemini-3.0 Prozero-shot=true2026.0579.672.283.578.4
Audio-Flamingo3zero-shot=true2026.0579.67466.473.3
Audio Flamingo 3Params=7B2026.0279.5873.9566.3773.3
Audio Flamingo 3Model Category=Large Audio Reasoning Models2026.0279.5873.9566.3773.3
Step-Audio 2 miniParams=7B2026.0279.368.4468.1672.73
MPAR²-7B2026.0279.272.875.175.7
Covo-AudioParams=7B2026.0278.6876.0571.1775.3
Qwen3-OmniSize=30B-A3B2026.0178.6869.4669.3772.5
Qwen2.5-OmniParams=7B2026.0278.165.970.671.5
Qwen2.5-Omni-7BModel Category=Large Audio Language Models2026.0278.165.970.671.5
Qwen2.5-Omnizero-shot=true2026.0578.165.970.671.5
MPAR2Model Size=7B2026.0277.9871.374.4774.59
MPAR2Model Size=3B2026.0276.4265.1169.0470.17
GaMMA-8Bzero-shot=true, parameters=8B2026.0576.373.76471.3
DIFFA-2Size=8B2026.0176.2863.4769.0669.6
Kimi-Audiozero-shot=true2026.0575.766.862.268.2
Kimi-AudioParams=7B2026.0275.6866.7762.1668.2
Kimi-AudioSize=7B2026.0175.6866.7762.1668.2
Kimi-AudioModel Category=Large Audio Language Models2026.0275.6866.7762.1668.2
DIFFA-2Size=8B, factor-based parallel decoding=true2026.0175.3861.0868.4768.3
Gemini-2.5 Prozero-shot=true2026.0575.168.371.571.6
Gemini 2.5 Pro2026.0275.0868.2671.4771.6
Gemini 2.5 Flash2026.0273.2765.5776.5871.8
Gemini 2.5 FlashModel Category=Commercial Models2026.0273.2765.5776.5871.8
Qwen2.5-OmniSize=7B2026.0172.9761.6860.9665.2
Audio Flamingo 2Params=3.2B2026.0271.4770.9644.7462.4
MiniCPM-OSize=8B2026.0171.4765.5763.0666.7
Gemini 2.0 Flash2026.0271.1765.2775.0870.5
Gemini 2.0 Flash2026.0171.1765.2775.0870.5
Qwen2.5-OmniModel Size=3B2026.0270.2760.4859.1663.3
Qwen2.5-OmniModel Size=7B2026.0269.466.861.665.9
Audio-ReasonerModel Category=Large Audio Reasoning Models2026.0267.8769.1666.0767.7
Qwen2-Audiozero-shot=true2026.0567.356.355.359.6
Qwen2-Audio-InstructParams=8.4B2026.0267.2756.2955.2659.6
Qwen2-AudioSize=8B2026.0167.2756.2955.2659.6
Qwen2-Audio-InstructModel Category=Large Audio Language Models2026.0267.2756.2955.2659.6
Phi-4-multimodalSize=8B2026.0165.4764.3767.2765.7
GPT-4o Audiozero-shot=true2026.0564.656.366.760.8
GPT-4o Audio2026.0264.5656.2966.6762.5
GPT-4o-Audio2026.0164.5656.2966.6762.5
GPT-4o AudioModel Category=Commercial Models2026.0264.5656.2966.6762.5
Qwen-Audio-ChatSize=8B2026.0155.254430.0343.1
GPT-4o mini Audio2026.0250.7539.2269.0753
DIFFASize=8B2026.0146.2543.4159.4649.71
SalmonnSize=7B2026.0141.1437.1326.4334.9
SALMONNzero-shot=true2026.0541.137.126.434.9
LTUSize=7B2026.0120.4215.9715.9217.44