HumanOmni-7B: benchmark results
Provider: Other.
Aggregate Rating 1465 ± 35, rank #3047 of 10916 rated models, from 49 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AVI-Bench - Audio-Referenced Visual Hallucination | 86 | Task score (%) on audio-referenced visual hallucination (250 | 92 |
| SonicBench - Counting | 59 | Accuracy (%; number of sound events (PicoAudio); 100 recogni | 77.4 |
| AVI-Bench - Audio Multi-Instance Classification | 26.54 | Task score (%) on audio multi-instance classification (518 c | 76 |
| TraceAV-Bench - Spatiotemporal Localization | 35.68 | Accuracy (%) on the 227 Spatiotemporal Localization (SL) que | 71.4 |
| SonicBench - Texture | 86 | Accuracy (%; sound texture (ESC-50); 100 recognition and 100 | 68.8 |
| AVI-Bench - Audio-Visual Question Answering | 56.54 | Task score (%) on audio-visual question answering (469 binar | 64 |
| AVI-Bench - Audio-Visual Sensation QA | 15.46 | Task score (%) on primitive audio-visual sensation (388 mult | 64 |
| SonicBench - Brightness | 61 | Accuracy (%; spectral brightness (NSynth); 100 recognition a | 62.9 |
| SonicBench - Timbre | 57 | Accuracy (%; instrument timbre (NSynth); 100 recognition and | 62.9 |
| AVI-Bench - Reasoning | 49.94 | Stage score (%) for the reasoning stage: mean of the two hal | 60 |
| SonicBench | 55 | Accuracy (%; mean over 12 physical attributes of 100 recogni | 59.7 |
| AVI-Bench | 28.4 | Overall score (%): mean of the four stage averages (percepti | 56 |
Interactive version: theaggregate.ai/model?slug=humanomni-7b · How It Works · Data refreshed daily, snapshot 2026-11-08.