Text-to-Speech on Difficult-Word Benchmark German 500 words
0.3D-PERGRAFT
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| GRAFTInput Modality=Audio Hint2026.07 | 0.3 | 0.76 | 0.94 | 0.11 | 3.99 | 0.68 | |
| Qwen3-TTS-0.6BInput Modality=Text2026.07 | 0.49 | 1.13 | 0.9 | 0.07 | 4.03 | 0.69 | |
| XTTS-v2Input Modality=Text2026.07 | 0.5 | 1.25 | 0.9 | 0.11 | 3.18 | 0.69 | |
| Piper (ph.)Input Modality=Phonemes, Speaker Setting=Single-Speaker2026.07 | 0.55 | 1.19 | 0.87 | 0.13 | 3.36 | 0.14 | |
| MaskGCT (ph.)Input Modality=Phonemes2026.07 | 0.56 | 1.28 | 0.87 | 0.17 | 3.07 | 0.79 |