List your company

Human-likeness classification on S2S Turing (test)

96.48Accuracy (H-H)GPT-4o-Audio-Preview
7.861630.868353.87576.8817Feb 27, 2026

Evaluation results

MethodLinks
GPT-4o-Audio-Preview2026.0296.4827.080.6941.16
Baichuan-Audio2026.0281.6915.2812.536.28
Qwen2.5-Omni2026.0278.1723.6123.6141.63
Phi-4-Multimodal2026.0277.4614.5822.2237.91
Human Judgement2026.0270.2883.5763.8472.84
MiniCPM-o 2.62026.0267.6143.0629.8646.74
Gemini 2.5 pro2026.0257.7572.9257.6462.79
Voxtral Mini2026.0251.4150.6938.8946.98
Gemma 3n2026.0246.4844.4440.2843.72
Seallms-Audio2026.0211.2784.7272.9256.51