Text-to-Speech on Seed-TTS EN
1.39WERQwen3-Omni
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Qwen3-Omni#Param=30B2026.05 | 1.39 | — | — | — | |
| UAS-Audio2026.04 | 1.7 | — | — | — | |
| Audio-OmniModel Type=Unified Models2026.04 | 1.77 | — | — | — | |
| F5-TTSCFG Strategy=Base, Zero-shot=true2025.09 | 1.8 | — | 0.679 | — | |
| F5-TTSCFG Strategy=def_text, Zero-shot=true2025.09 | 1.8 | — | 0.69 | — | |
| F5-TTSModel Type=Specialized Models2026.04 | 1.83 | — | — | — | |
| Minimax-SpeechZero-shot=true2025.09 | 1.9 | — | 0.738 | — | |
| GLM-4-Voice#Param=9B2026.05 | 2.1 | — | — | — | |
| Seed-TTSZero-shot=true2025.09 | 2.2 | — | 0.762 | — | |
| CosyVoice 3-1.5BZero-shot=true2025.09 | 2.2 | — | 0.72 | — | |
| VITA-QINYU-4B#Param=4B2026.05 | 2.2 | — | — | — | |
| Seed-TTS#Param=-2026.05 | 2.25 | — | — | — | |
| VITA-QINYU-8B#Param=8B2026.05 | 2.29 | — | — | — | |
| Qwen2.5-Omni2026.04 | 2.3 | — | — | — | |
| Qwen2.5-Omni#Param=7B2026.05 | 2.33 | — | — | — | |
| CosyVoice 2CFG Strategy=Base, Zero-shot=true2025.09 | 2.4 | — | 0.66 | — | |
| CosyVoice3Model Type=Specialized Models2026.04 | 2.46 | — | — | — | |
| CosyVoice2#Param=0.5B2026.05 | 2.57 | — | — | — | |
| CosyVoice2Model Category=Audio, Evaluation Protocol=audio-only2026.05 | 2.57 | — | — | 65.2 | |
| CosyVoice 2CFG Strategy=input_text, Zero-shot=true2025.09 | 2.6 | — | 0.666 | — | |
| MaskGCTModel Type=Specialized Models2026.04 | 2.62 | — | — | — | |
| Qwen2.5-OmniModel Category=Audio, Evaluation Protocol=audio-only2026.05 | 2.72 | — | — | 63.2 | |
| Qwen2.5-OmniParameters=7B2026.02 | 3.1 | 3.62 | — | — | |
| Step-Audio2-mini2026.04 | 3.2 | — | — | — | |
| UniAudio 2.0Parameters=3B2026.02 | 3.63 | 3.8 | — | — | |
| NAVAModel Category=Audio-Video, Evaluation Protocol=audio-only2026.05 | 4.2 | — | — | 66.7 | |
| CosyVoiceModel Category=Audio, Evaluation Protocol=audio-only2026.05 | 4.29 | — | — | 60.9 | |
| Ming-OmniModel Type=Unified Models2026.04 | 4.31 | — | — | — | |
| Baichuan-Audio-Instruct2026.04 | 4.7 | — | — | — | |
| MiMo-Audio-InParameters=7B2026.02 | 4.74 | 3.27 | — | — | |
| Unified-IO2Model Type=Unified Models2026.04 | 21.63 | — | — | — | |
| DreamID-OmniModel Category=Audio-Video, Evaluation Protocol=audio-only2026.05 | 31.76 | — | — | 35.7 |