# Speech-to-Speech reasoning on Kimi-Audio-Evalkit s2s Source: [Speech-to-Speech reasoning on Kimi-Audio-Evalkit s2s]() Collection index: [llms.txt]() Data updated: 2026-07-01T07:32:11.722Z Evaluation of model performance on Speech-to-Speech reasoning tasks using the Kimi-Audio-Evalkit s2s dataset, reporting metrics across various reasoning and quality dimensions. 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-Speech reasoning | | Dataset (including any stated split) | Kimi-Audio-Evalkit s2s | | Metric | Llama Q Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 17 of 17 | | 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 - [Llama Q Score](): Higher is better; unit: PERCENT. - [Web Q Score](): Higher is better; unit: PERCENT. - [Trivia Score](): Higher is better; unit: PERCENT. - [Alpaca Score](): Higher is better; unit: SCALAR. - [Common Score](): Higher is better; unit: SCALAR. - [IFEval Score](): Higher is better; unit: SCALAR. - [SD-QA Score](): Higher is better; unit: PERCENT. - [AdvB Score](): Higher is better; unit: SCALAR. - [Overall Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Llama Q Score | Result date | Report paper | | --- | --- | --- | --- | | Gemini 2.5-Flash | 78 | Not recorded | 6a449eeee0224bd4f4992db8 | | FlexiSLM-7B-Stage2 | 75 | Not recorded | 6a449eeee0224bd4f4992db8 | | FlexiSLM-7B-Stage3 | 75 | Not recorded | 6a449eeee0224bd4f4992db8 | | Fun-Audio-Chat-8B | 74 | Not recorded | 6a449eeee0224bd4f4992db8 | | FlexiSLM-7B-Stage2 | 74 | Not recorded | 6a449eeee0224bd4f4992db8 | | FlexiSLM-7B-Stage3 | 74 | Not recorded | 6a449eeee0224bd4f4992db8 | | FlexiSLM-7B-Stage3 | 73 | Not recorded | 6a449eeee0224bd4f4992db8 | | Qwen3-Omni-30B-A3B | 72 | Not recorded | 6a449eeee0224bd4f4992db8 | | Qwen2.5-Omni-7B | 72 | Not recorded | 6a449eeee0224bd4f4992db8 | | Mimo-Audio-7B | 71 | Not recorded | 6a449eeee0224bd4f4992db8 | | FlexiSLM-7B-Stage2 | 71 | Not recorded | 6a449eeee0224bd4f4992db8 | | FlexiSLM-7B-Stage3 | 71 | Not recorded | 6a449eeee0224bd4f4992db8 | | FlexiSLM-7B-Stage2 | 70 | Not recorded | 6a449eeee0224bd4f4992db8 | | Kimi-Audio-7B | 69 | Not recorded | 6a449eeee0224bd4f4992db8 | | FlexiSLM-7B-Stage3 | 67 | Not recorded | 6a449eeee0224bd4f4992db8 | | FlexiSLM-7B-Stage3 | 64 | Not recorded | 6a449eeee0224bd4f4992db8 | | GLM-4-Voice-9B | 62 | Not recorded | 6a449eeee0224bd4f4992db8 | ### Gemini 2.5-Flash Llama Q Score: 78. Result ID: 6a44c27a73d097d845be74b1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction. ### FlexiSLM-7B-Stage2 Llama Q Score: 75. Result ID: 6a44c27a73d097d845be74b8. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"In FR":"12.5","Out FR":"12.5"}. This record does not assert independent reproduction. ### FlexiSLM-7B-Stage3 Llama Q Score: 75. Result ID: 6a44c27a73d097d845be74bc. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"In FR":"12.5","Out FR":"12.5"}. This record does not assert independent reproduction. ### Fun-Audio-Chat-8B Llama Q Score: 74. Result ID: 6a44c27a73d097d845be74b4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"In FR":"5.0","Out FR":"5.0"}. This record does not assert independent reproduction. ### FlexiSLM-7B-Stage2 Llama Q Score: 74. Result ID: 6a44c27a73d097d845be74ba. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"In FR":"6.25","Out FR":"12.5"}. This record does not assert independent reproduction. ### FlexiSLM-7B-Stage3 Llama Q Score: 74. Result ID: 6a44c27a73d097d845be74bd. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"In FR":"12.5","Out FR":"6.25"}. This record does not assert independent reproduction. ### FlexiSLM-7B-Stage3 Llama Q Score: 73. Result ID: 6a44c27a73d097d845be74be. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"In FR":"6.25","Out FR":"12.5"}. This record does not assert independent reproduction. ### Qwen3-Omni-30B-A3B Llama Q Score: 72. Result ID: 6a44c27a73d097d845be74b2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"In FR":"12.5","Out FR":"12.5"}. This record does not assert independent reproduction. ### Qwen2.5-Omni-7B Llama Q Score: 72. Result ID: 6a44c27a73d097d845be74b7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"In FR":"25","Out FR":"50"}. This record does not assert independent reproduction. ### Mimo-Audio-7B Llama Q Score: 71. Result ID: 6a44c27a73d097d845be74b5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"In FR":"6.25","Out FR":"6.25"}. This record does not assert independent reproduction. ### FlexiSLM-7B-Stage2 Llama Q Score: 71. Result ID: 6a44c27a73d097d845be74bb. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"In FR":"6.25","Out FR":"6.25"}. This record does not assert independent reproduction. ### FlexiSLM-7B-Stage3 Llama Q Score: 71. Result ID: 6a44c27a73d097d845be74bf. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"In FR":"6.25","Out FR":"6.25"}. This record does not assert independent reproduction. ### FlexiSLM-7B-Stage2 Llama Q Score: 70. Result ID: 6a44c27a73d097d845be74b9. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"In FR":"12.5","Out FR":"6.25"}. This record does not assert independent reproduction. ### Kimi-Audio-7B Llama Q Score: 69. Result ID: 6a44c27a73d097d845be74b6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"In FR":"12.5","Out FR":"12.5"}. This record does not assert independent reproduction. ### FlexiSLM-7B-Stage3 Llama Q Score: 67. Result ID: 6a44c27a73d097d845be74c0. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"In FR":"5.0","Out FR":"5.0"}. This record does not assert independent reproduction. ### FlexiSLM-7B-Stage3 Llama Q Score: 64. Result ID: 6a44c27a73d097d845be74c1. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"In FR":"4.0","Out FR":"4.0"}. This record does not assert independent reproduction. ### GLM-4-Voice-9B Llama Q Score: 62. Result ID: 6a44c27a73d097d845be74b3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"In FR":"12.5","Out FR":"12.5"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.