Speech-to-Text Reply Quality on VoiceBench AlpacaEval
4.78Quality ScoreGPT-4o-Audio
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-4o-Audio2026.05 | 4.78 | |
| Listen-Write-Speak2026.05 | 4.72 | |
| VITA-1.52026.05 | 4.21 | |
| Step-Audio2026.05 | 4.13 | |
| Freeze-Omni2026.05 | 4.03 | |
| GLM-4-Voice2026.05 | 3.97 |