AudioProcessBench (Zero-shot): leaderboard

Metric: Overall score (0-100): mean of the step-correctness, error-type and chain-aggregation averages for an audio model prompted as a step-level critic without examples; 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; zero-shot critic prompt, temperature 0.7; random 37.2; higher is better. Source: arxiv.org. Saturation forecast: Around 2029. 11 models tracked.

Top models

#ModelScore
1Gemini 3 Flash64.3
2Phi-4 Multimodal Instruct47.3
3Gemma 3n E4B44.5
4Gemma 3n E2B44.2
5Qwen2.5-Omni-7B41.1
6Gemma 4 E2B40.3
7Gemma 4 E4B38.1

Interactive version: theaggregate.ai/benchmark?slug=audioprocessbench-zero-shot · How It Works · Data refreshed daily, snapshot 2026-09-29.