# Zero-shot Text-to-Speech on EARS (unseen speakers)

Source: [Zero-shot Text-to-Speech on EARS (unseen speakers)](<https://www.sota2.com/research/sota/zero-shot-text-to-speech-on-ears-unseen-speakers>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-03-17T10:14:44.469Z

Evaluation of Zero-shot Text-to-Speech performance specifically on the unseen speakers subset of the EARS dataset, utilizing metrics related to nonverbal and verbal quality/accuracy (e.g., WER, NV-Acc, EECS/SECS scores).

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Zero-shot Text-to-Speech |
| Dataset (including any stated split) | EARS (unseen speakers) |
| Metric | WER |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 7 of 7 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [WER](<https://www.sota2.com/research/sota/zero-shot-text-to-speech-on-ears-unseen-speakers.md?metric=WER>): Lower is better; unit: SCALAR.

- [NV-Acc](<https://www.sota2.com/research/sota/zero-shot-text-to-speech-on-ears-unseen-speakers.md?metric=NV-Acc>): Higher is better; unit: PERCENT.

- [NV-Sim](<https://www.sota2.com/research/sota/zero-shot-text-to-speech-on-ears-unseen-speakers.md?metric=NV-Sim>): Higher is better; unit: PERCENT.

- [Nonverbal EECS](<https://www.sota2.com/research/sota/zero-shot-text-to-speech-on-ears-unseen-speakers.md?metric=Nonverbal%20EECS>): Higher is better; unit: PERCENT.

- [Nonverbal SECS](<https://www.sota2.com/research/sota/zero-shot-text-to-speech-on-ears-unseen-speakers.md?metric=Nonverbal%20SECS>): Higher is better; unit: PERCENT.

- [Verbal EECS](<https://www.sota2.com/research/sota/zero-shot-text-to-speech-on-ears-unseen-speakers.md?metric=Verbal%20EECS>): Higher is better; unit: PERCENT.

- [Verbal SECS](<https://www.sota2.com/research/sota/zero-shot-text-to-speech-on-ears-unseen-speakers.md?metric=Verbal%20SECS>): Higher is better; unit: PERCENT.

## Recorded results

| Method | WER | Result date | Report paper |
| --- | --- | --- | --- |
| Fun-CosyVoice3-0.5B | 1.65 | Not recorded | 69b8e0180636e5bf4afbeaed |
| CosyVoice2-0.5B | 1.97 | Not recorded | 69b8e0180636e5bf4afbeaed |
| Augmented GT | 2.93 | Not recorded | 69b8e0180636e5bf4afbeaed |
| Affectron-330M | 8.31 | Not recorded | 69b8e0180636e5bf4afbeaed |
| VoiceCraft-330M | 8.89 | Not recorded | 69b8e0180636e5bf4afbeaed |
| VoiceCraft-330M | 10.5 | Not recorded | 69b8e0180636e5bf4afbeaed |
| Dia-1.6B | 11.5 | Not recorded | 69b8e0180636e5bf4afbeaed |

### Fun-CosyVoice3-0.5B

WER: 1.65. Result ID: 69b9298347f34ce924a26c4a.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e0180636e5bf4afbeaed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78dd1e209131cea7ae5b.md>)

Recorded evaluation configuration: {"Evaluation Protocol":"Pre-trained"}. This record does not assert independent reproduction.

### CosyVoice2-0.5B

WER: 1.97. Result ID: 69b9298347f34ce924a26c49.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e0180636e5bf4afbeaed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa76591e209131cea6c851.md>)

Recorded evaluation configuration: {"Evaluation Protocol":"Pre-trained"}. This record does not assert independent reproduction.

### Augmented GT

WER: 2.93. Result ID: 69b9298347f34ce924a26c47.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e0180636e5bf4afbeaed.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Affectron-330M

WER: 8.31. Result ID: 69b9298347f34ce924a26c4d.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e0180636e5bf4afbeaed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69b8e0180636e5bf4afbeaed.md>)

Recorded evaluation configuration: {"Evaluation Protocol":"Fine-tuned"}. This record does not assert independent reproduction.

### VoiceCraft-330M

WER: 8.89. Result ID: 69b9298347f34ce924a26c48.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e0180636e5bf4afbeaed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73a61e209131cea5daf3.md>)

Recorded evaluation configuration: {"Evaluation Protocol":"Pre-trained"}. This record does not assert independent reproduction.

### VoiceCraft-330M

WER: 10.5. Result ID: 69b9298347f34ce924a26c4c.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e0180636e5bf4afbeaed.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73a61e209131cea5daf3.md>)

Recorded evaluation configuration: {"Evaluation Protocol":"Fine-tuned"}. This record does not assert independent reproduction.

### Dia-1.6B

WER: 11.5. Result ID: 69b9298347f34ce924a26c4b.

[Reporting paper](<https://www.sota2.com/research/paper/69b8e0180636e5bf4afbeaed.md>)

Recorded evaluation configuration: {"Evaluation Protocol":"Pre-trained"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.