# Speech Quality Evaluation on QA Evaluation Suite Speech Output Source: [Speech Quality Evaluation on QA Evaluation Suite Speech Output]() Collection index: [llms.txt]() Data updated: 2026-03-11T07:26:40.207Z Evaluation of speech quality using Mean Opinion Score (MOS) and ASR Word Error Rate (ASR-WER) on the QA Evaluation Suite Speech Output dataset. 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 Quality Evaluation | | Dataset (including any stated split) | QA Evaluation Suite Speech Output | | Metric | MOS | | Unit | SCALAR | | Direction | Higher is better | | Results loaded | 16 of 16 | | 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 - [MOS](): Higher is better; unit: SCALAR. - [ASR-WER](): Lower is better; unit: PERCENT. ## Recorded results | Method | MOS | Result date | Report paper | | --- | --- | --- | --- | | Step-Audio-2 | 3.71 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | SPAR-K | 3.668 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | SPAR-K | 3.65 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | SPAR-K | 3.643 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | Step-Audio-2 | 3.058 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | GLM-4-Voice | 2.982 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | SPAR-K | 2.982 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | SPAR-K | 2.981 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | SPAR-K | 2.979 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | SPAR-K | 2.95 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | SPAR-K | 2.949 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | SPAR-K | 2.942 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | SPAR-K | 2.909 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | GLM-4-Voice | 2.866 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | GLM-4-Voice | 2.662 | Not recorded | 69b0f6fe2d603e6de3e45d1b | | Step-Audio-2 | 1.651 | Not recorded | 69b0f6fe2d603e6de3e45d1b | ### Step-Audio-2 MOS: 3.71. Result ID: 69b1192f6dbb7075b95b2071. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Step-Audio-2","Text EE":"-","Speech EE":"-"}. This record does not assert independent reproduction. ### SPAR-K MOS: 3.668. Result ID: 69b1192f6dbb7075b95b2076. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Step-Audio-2","Text EE":"-","Speech EE":"Triple (22)"}. This record does not assert independent reproduction. ### SPAR-K MOS: 3.65. Result ID: 69b1192f6dbb7075b95b2075. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Step-Audio-2","Text EE":"-","Speech EE":"Odd (22)"}. This record does not assert independent reproduction. ### SPAR-K MOS: 3.643. Result ID: 69b1192f6dbb7075b95b2074. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Step-Audio-2","Text EE":"-","Speech EE":"Even (22)"}. This record does not assert independent reproduction. ### Step-Audio-2 MOS: 3.058. Result ID: 69b1192f6dbb7075b95b2072. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Step-Audio-2","Text EE":"-","Speech EE":"Fixed (25)"}. This record does not assert independent reproduction. ### GLM-4-Voice MOS: 2.982. Result ID: 69b1192f6dbb7075b95b2077. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GLM-4-Voice","Text EE":"-","Speech EE":"-"}. This record does not assert independent reproduction. ### SPAR-K MOS: 2.982. Result ID: 69b1192f6dbb7075b95b207b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GLM-4-Voice","Text EE":"-","Speech EE":"Odd (36)"}. This record does not assert independent reproduction. ### SPAR-K MOS: 2.981. Result ID: 69b1192f6dbb7075b95b207d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GLM-4-Voice","Text EE":"-","Speech EE":"Even (33)"}. This record does not assert independent reproduction. ### SPAR-K MOS: 2.979. Result ID: 69b1192f6dbb7075b95b207f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GLM-4-Voice","Text EE":"Even (36)","Speech EE":"-"}. This record does not assert independent reproduction. ### SPAR-K MOS: 2.95. Result ID: 69b1192f6dbb7075b95b207a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GLM-4-Voice","Text EE":"-","Speech EE":"Even (36)"}. This record does not assert independent reproduction. ### SPAR-K MOS: 2.949. Result ID: 69b1192f6dbb7075b95b2080. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GLM-4-Voice","Text EE":"Conf (36)","Speech EE":"Even(36)"}. This record does not assert independent reproduction. ### SPAR-K MOS: 2.942. Result ID: 69b1192f6dbb7075b95b207e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GLM-4-Voice","Text EE":"-","Speech EE":"Even (35)"}. This record does not assert independent reproduction. ### SPAR-K MOS: 2.909. Result ID: 69b1192f6dbb7075b95b207c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GLM-4-Voice","Text EE":"-","Speech EE":"Triple (37)"}. This record does not assert independent reproduction. ### GLM-4-Voice MOS: 2.866. Result ID: 69b1192f6dbb7075b95b2079. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GLM-4-Voice","Text EE":"-","Speech EE":"Conf (36)"}. This record does not assert independent reproduction. ### GLM-4-Voice MOS: 2.662. Result ID: 69b1192f6dbb7075b95b2078. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GLM-4-Voice","Text EE":"-","Speech EE":"Fixed (38)"}. This record does not assert independent reproduction. ### Step-Audio-2 MOS: 1.651. Result ID: 69b1192f6dbb7075b95b2073. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Step-Audio-2","Text EE":"-","Speech EE":"Conf (26)"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.