SEA-SpeechBench - Temporal Localization (0-30 s): leaderboard
Metric: Span-overlap F1 (%; predict the start and end time at which a queried phrase is spoken in a recording of 0-30 seconds; English instruction prompt; blank where the model cannot take audio of that length). Source: arxiv.org. Saturation forecast: Around 2035. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen2.5-Omni-7B | 35.74 |
| 2 | Qwen2-Audio-7B-Instruct | 33.3 |
| 3 | GPT-4o Audio | 26.47 |
| 4 | Voxtral-Mini-3B-2507 | 17.85 |
| 5 | Gemma 3n E4B (IT) | 13.25 |
| 6 | Phi-4 Multimodal Instruct | 12.97 |
| 7 | gemma-3n-E2B-it | 11.82 |
| 8 | Gemini 2.5 Flash | 11.66 |
| 9 | SeaLLMs-Audio-7B | 11.57 |
| 10 | Qwen3 Omni 30B A3B Instruct | 11.4 |
Interactive version: theaggregate.ai/benchmark?slug=sea-speechbench-temporal-localization-0-30-s · How It Works · Data refreshed daily, snapshot 2026-09-26.