# Audio Understanding on SpeechIO

Source: [Audio Understanding on SpeechIO](<https://www.sota2.com/research/sota/audio-understanding-on-speechio>)

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

Data updated: 2026-03-27T16:49:03.898Z

Evaluation of general audio understanding performance on the SpeechIO dataset, measured using Error Rate.

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 | Audio Understanding |
| Dataset (including any stated split) | SpeechIO |
| Metric | Error Rate |
| Unit | PERCENT |
| Direction | Lower is better |
| Results loaded | 4 of 4 |
| 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

- [Error Rate](<https://www.sota2.com/research/sota/audio-understanding-on-speechio.md?metric=Error%20Rate>): Lower is better; unit: PERCENT.

## Recorded results

| Method | Error Rate | Result date | Report paper |
| --- | --- | --- | --- |
| Qwen3 Omni | 2.18 | Not recorded | 6902f0f23ef7ccd74e5e4274 |
| Kimi Audio | 2.23 | Not recorded | 6902f0f23ef7ccd74e5e4274 |
| Ming-Flash-Omni | 2.49 | Not recorded | 6902f0f23ef7ccd74e5e4274 |
| Qwen2 Audio | 3.01 | Not recorded | 6902f0f23ef7ccd74e5e4274 |

### Qwen3 Omni

Error Rate: 2.18. Result ID: 69c6b48c47f34ce924b11a41.

[Reporting paper](<https://www.sota2.com/research/paper/6902f0f23ef7ccd74e5e4274.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7b881e209131cea88f3d.md>)

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

### Kimi Audio

Error Rate: 2.23. Result ID: 69c6b48c47f34ce924b11a43.

[Reporting paper](<https://www.sota2.com/research/paper/6902f0f23ef7ccd74e5e4274.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78431e209131cea77803.md>)

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

### Ming-Flash-Omni

Error Rate: 2.49. Result ID: 69c6b48c47f34ce924b11a40.

[Reporting paper](<https://www.sota2.com/research/paper/6902f0f23ef7ccd74e5e4274.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6902f0f23ef7ccd74e5e4274.md>)

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

### Qwen2 Audio

Error Rate: 3.01. Result ID: 69c6b48c47f34ce924b11a42.

[Reporting paper](<https://www.sota2.com/research/paper/6902f0f23ef7ccd74e5e4274.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.