# Spoken Question Answering on TriviaQA (speech score, text score)

Source: [Spoken Question Answering on TriviaQA (speech score, text score)](<https://www.sota2.com/research/sota/spoken-question-answering-on-triviaqa-speech-score-text-score>)

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

Data updated: 2026-05-12T15:02:43.515Z

Evaluation of model performance on the TriviaQA dataset for Spoken Question Answering, reporting custom speech and text derived 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 | Spoken Question Answering |
| Dataset (including any stated split) | TriviaQA |
| Metric | Text Score |
| Unit | SCALAR |
| Direction | Higher 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

- [Text Score](<https://www.sota2.com/research/sota/spoken-question-answering-on-triviaqa-speech-score-text-score.md?metric=Text%20Score>): Higher is better; unit: SCALAR.

- [Speech Score](<https://www.sota2.com/research/sota/spoken-question-answering-on-triviaqa-speech-score-text-score.md?metric=Speech%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Text Score | Result date | Report paper |
| --- | --- | --- | --- |
| Freeze-Omni | 45.1 | Not recorded | 6a02b49bdbc836b87d824e14 |
| GLM-4-Voice | 39.1 | Not recorded | 6a02b49bdbc836b87d824e14 |
| Moshi | 25.6 | Not recorded | 6a02b49bdbc836b87d824e14 |
| CF-Duplex | 19.6 | Not recorded | 6a02b49bdbc836b87d824e14 |
| Speech-GPT | 14.8 | Not recorded | 6a02b49bdbc836b87d824e14 |
| XA-Duplex | 8.2 | Not recorded | 6a02b49bdbc836b87d824e14 |

### Freeze-Omni

Text Score: 45.1. Result ID: 6a03407b47f34ce924e83ccf.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b49bdbc836b87d824e14.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa760a1e209131cea6b0d4.md>)

Recorded evaluation configuration: {"Backbone LLM":"Qwen2-7B-Instruct","Full-Duplex":true}. This record does not assert independent reproduction.

### GLM-4-Voice

Text Score: 39.1. Result ID: 6a03407b47f34ce924e83cce.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b49bdbc836b87d824e14.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75f91e209131cea6abba.md>)

Recorded evaluation configuration: {"Backbone LLM":"GLM-4-9B-Base","Full-Duplex":false}. This record does not assert independent reproduction.

### Moshi

Text Score: 25.6. Result ID: 6a03407b47f34ce924e83cd0.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b49bdbc836b87d824e14.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75001e209131cea652b7.md>)

Recorded evaluation configuration: {"Backbone LLM":"Helium (7B)","Full-Duplex":true}. This record does not assert independent reproduction.

### CF-Duplex

Text Score: 19.6. Result ID: 6a03407b47f34ce924e83cd2.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b49bdbc836b87d824e14.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a02b49bdbc836b87d824e14.md>)

Recorded evaluation configuration: {"Backbone LLM":"Qwen3-1.7B","Full-Duplex":true}. This record does not assert independent reproduction.

### Speech-GPT

Text Score: 14.8. Result ID: 6a03407b47f34ce924e83ccd.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b49bdbc836b87d824e14.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f701e209131cea45919.md>)

Recorded evaluation configuration: {"Backbone LLM":"LLaMA-13B","Full-Duplex":false}. This record does not assert independent reproduction.

### XA-Duplex

Text Score: 8.2. Result ID: 6a03407b47f34ce924e83cd3.

[Reporting paper](<https://www.sota2.com/research/paper/6a02b49bdbc836b87d824e14.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a02b49bdbc836b87d824e14.md>)

Recorded evaluation configuration: {"Backbone LLM":"Qwen3-1.7B","Full-Duplex":true}. This record does not assert independent reproduction.

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