List your company

General Audio Understanding on VoiceBench

86.43Overall ScoreGPT-4o-Audio
25.090841.015456.9472.8646Jan 30, 2026Feb 20, 2026Mar 14, 2026Apr 5, 2026Apr 26, 2026May 18, 2026Jun 9, 2026

Evaluation results

MethodLinks
GPT-4o-Audio2026.0186.434.784.4975.580.2589.2376.0298.65
Kimi-Audio2026.0176.934.463.9763.1262.1783.5261.1100
Qwen-2.5-Omni2026.0174.044.53.8456.461.780.953.599.2
GRPOInference Setting=Scaffold-free2026.0669.9———————
BaselineInference Setting=Scaffold-free2026.0668.98———————
ParaBridgeInference Setting=Scaffold-free2026.0668.63———————
RFTInference Setting=Scaffold-free2026.0668.31———————
Phi-4-multimodal2026.0163.693.813.8239.7842.1965.9345.35100
DIFFA-22026.0159.633.863.6740.7838.1361.5340.7685.58
GLM-4-Voice2026.0155.993.973.4236.9839.7553.4125.9288.08
DiVA2026.0155.73.673.5457.0625.7625.4939.1698.27
Qwen2-Audio2026.0155.343.743.4335.7235.7249.4526.3396.73
Freeze-Omni2026.0154.724.033.4653.4528.1430.9823.497.3
Step-Audio2026.0149.774.133.0944.2128.3333.8527.9669.62
DIFFA2026.0148.223.782.9634.4529.5735.626.5676.54
LLaMA-Omni2026.0137.53.73.4639.6925.9327.4714.8711.35
VITA2026.0134.683.382.1527.9425.729.0122.8226.73
Slam-Omni2026.0133.841.91.794.1626.0625.2713.3894.23
Mini-Omni22026.0131.322.322.189.3124.2726.5911.5657.5
Moshi2026.0127.452.011.615.6424.0425.9310.1244.23
AF-Next-InstructTask=VoiceBench2026.04—4.433.96——80.9—98.84
Audio Flamingo 3Task=VoiceBench2026.04—4.193.4——66.81—98.26
Qwen2.5-OTask=VoiceBench2026.04—4.333.84——79.12—99.62