GPT-4o Audio (Preview): benchmark results
Provider: OpenAI. Released 2024-10-17. Access: API.
Aggregate Rating 1638 ± 48, rank #708 of 10916 rated models, from 8 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| WavBench - Explicit Acoustic Generation | 79.23 | Accuracy (%) of generated speech against the requested paral | 100 |
| HalluAudio - Speech | 45.51 | Average classification accuracy (%) over the ten speech task | 90 |
| HalluAudio - Music | 58.15 | Average classification accuracy (%) over the nine music task | 88.9 |
| HalluAudio - Environmental Sound | 71 | Average classification accuracy (%) over the seven environme | 62.5 |
| AGL1K | 1996.15 | Geo-score (0-5000) | 60 |
| AGL1K - Speech Distance Error | 2841.03 | Mean distance error on clips with speech (km) | 60 |
| WavBench - Explicit Acoustic Understanding | 48.7 | Accuracy (%) of the model's predictions against ground-truth | 25 |
| AGL1K - Non-Speech Distance Error | 7069.09 | Mean distance error on clips without speech (km) | 20 |
Interactive version: theaggregate.ai/model?slug=gpt-4o-audio-preview · How It Works · Data refreshed daily, snapshot 2026-11-08.