# Speech-to-Text Reply Quality on VoiceBench AlpacaEval

Source: [Speech-to-Text Reply Quality on VoiceBench AlpacaEval](<https://www.sota2.com/research/sota/speech-to-text-reply-quality-on-voicebench-alpacaeval>)

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

Data updated: 2026-06-09T06:22:43.282Z

Evaluation of model performance on the VoiceBench AlpacaEval dataset, focusing specifically on the quality assessment of generated replies in a Speech-to-Text interaction.

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 | Speech-to-Text Reply Quality |
| Dataset (including any stated split) | VoiceBench AlpacaEval |
| Metric | Quality Score |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 6 of 6 |
| 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

- [Quality Score](<https://www.sota2.com/research/sota/speech-to-text-reply-quality-on-voicebench-alpacaeval.md?metric=Quality%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Quality Score | Result date | Report paper |
| --- | --- | --- | --- |
| GPT-4o-Audio | 4.78 | Not recorded | 6a279e0890966ade919bf642 |
| Listen-Write-Speak | 4.72 | Not recorded | 6a279e0890966ade919bf642 |
| VITA-1.5 | 4.21 | Not recorded | 6a279e0890966ade919bf642 |
| Step-Audio | 4.13 | Not recorded | 6a279e0890966ade919bf642 |
| Freeze-Omni | 4.03 | Not recorded | 6a279e0890966ade919bf642 |
| GLM-4-Voice | 3.97 | Not recorded | 6a279e0890966ade919bf642 |

### GPT-4o-Audio

Quality Score: 4.78. Result ID: 6a27b03cfbc373592c23a719.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e0890966ade919bf642.md>)

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

### Listen-Write-Speak

Quality Score: 4.72. Result ID: 6a27b03cfbc373592c23a71e.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e0890966ade919bf642.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a279e0890966ade919bf642.md>)

[Implementation](<https://github.com/zly-idleness/lws_demo>)

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

### VITA-1.5

Quality Score: 4.21. Result ID: 6a27b03cfbc373592c23a71a.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e0890966ade919bf642.md>)

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

### Step-Audio

Quality Score: 4.13. Result ID: 6a27b03cfbc373592c23a71b.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e0890966ade919bf642.md>)

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

### Freeze-Omni

Quality Score: 4.03. Result ID: 6a27b03cfbc373592c23a71c.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e0890966ade919bf642.md>)

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

### GLM-4-Voice

Quality Score: 3.97. Result ID: 6a27b03cfbc373592c23a71d.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e0890966ade919bf642.md>)

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

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