List your company
ResearchFormerly WizWand

Text-to-Speech on Seed-TTS Seed-EN (test)

0.0147WERARCHI-TTS
-0.025640.2466550.518950.791245Feb 5, 2026Feb 23, 2026Mar 14, 2026Apr 2, 2026Apr 20, 2026May 9, 2026May 28, 2026
Updated 5mo ago

Evaluation results

MethodLinks
ARCHI-TTS2026.020.01470.68
DiTAR2026.020.01690.74
SemaVoiceType=C-AR, Params=1.5B, #Hours=150K2026.050.01710.694
Seed-TTS_DiT2026.020.01730.79
F5-TTS2026.020.01830.67
VoxCPMType=C-AR, Params=0.5B, #Hours=1.8M2026.050.01850.729
SemaVoice-EmiliaType=C-AR, Params=1.5B, #Hours=100K2026.050.01910.657
OpenAudio-s1Type=D-AR, Params=0.5B, #Hours=2.0M2026.050.01940.55
FireRedTTS-2Type=D-AR, Params=-, #Hours=1.4M2026.050.01950.665
SparkTTSType=D-AR, Params=0.5B, #Hours=100K2026.050.01980.573
F5-TTSType=C-NAR, Params=0.3B, #Hours=100K2026.050.020.647
Ground Truth2026.020.02060.73
VAE Resynthesized2026.020.02060.7
Vocos Resynthesized2026.020.02090.7
Ground TruthType=-, Params=-, #Hours=-2026.050.02140.734
E2 TTS2026.020.02190.71
IndexTTS 2Type=D-AR+C-NAR, Params=1.5B, #Hours=55K2026.050.02230.706
VoxCPM-EmiliaType=C-AR, Params=0.5B, #Hours=100K2026.050.02340.681
HiggsAudio-v2Type=D-AR, Params=3.0B, #Hours=10M2026.050.02440.677
CosyVoice 22026.020.02570.65
CosyVoice2Type=D-AR+C-NAR, Params=0.5B, #Hours=170K2026.050.02570.659
MaskGCTType=D-NAR, Params=1.0B, #Hours=100K2026.050.02620.717
MaskGCT2026.020.02620.717
Qwen2.5-OmniType=MLLM, Params=7.0B, #Hours=-2026.050.02720.632
VibeVoiceType=C-AR, Params=1.5B, #Hours=-2026.050.03040.689
FireRedTTS2026.020.03820.46
FireRedTTSType=D-AR+C-NAR, Params=0.5B, #Hours=248K2026.050.03820.46
CosyVoiceType=D-AR+C-NAR, Params=0.3B, #Hours=170K2026.050.04290.609
HoliTok-UniteModeling=Unified spoken language modeling2026.050.0720.55
HoliTok-BaseModeling=Unified spoken language modeling2026.050.27850.52
MingTok-AudioModeling=Unified spoken language modeling2026.050.51060.42
Semantic-VAEModeling=Unified spoken language modeling2026.051.02320.47