AudioProcessBench - Best-of-N Selection: leaderboard
AudioProcessBench (2026) tests audio and omni-modal models as critics of audio-grounded reasoning: given an audio question and a model-written reasoning chain split into steps, the critic marks each step correct or wrong; 3,872 annotated chains from MMAR, MMSU and MMAU-Pro questions. This board uses the step judgments of the critic to pick the best of several chains for the same question and reports final-answer accuracy in percent; higher is better.
Metric: Final-answer accuracy (%) when the critic's summed step scores pick one chain among several sampled for the same audio question; 3,872 reasoning traces (23,497 steps, 41% erroneous) from six audio and omni models on MMAR, MMSU and MMAU-Pro questions, steps labelled by two LLM annotators with human review; few-shot critic prompt with three annotated example chains, temperature 0.7; random 47.5; higher is better. Source: arxiv.org. Saturation forecast: Around 2030. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash | 59.4 | #164 |
| 2 | Phi-4 Multimodal Instruct | 52.9 | #3882 |
| 3 | Gemma 4 E2B | 52.5 | #3892 |
| 4 | Gemma 4 E4B | 51.6 | #2214 |
| 5 | Qwen2.5-Omni-7B | 51.1 | #1789 |
| 6 | Gemma 3n E4B | 50.8 | #3301 |
| 7 | Gemma 3n E2B | 50.3 | #6091 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=audioprocessbench-best-of-n-selection · How It Works · Data refreshed daily, snapshot 2026-11-13.