Audio-Visual Reasoning on OmniBench
63.6AccuracyQwen3-Omni
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-OmniReasoning=true, Model Size=30B, Zero-shot=true2026.04 | 63.6 | |
| AVRTReasoning=true, Model Size=7B, Zero-shot=true2026.04 | 57.1 | |
| AVRTReasoning=true, Model Size=3B, Zero-shot=true2026.04 | 56.3 | |
| Qwen2.5 OmniReasoning=false, Model Size=7B, Zero-shot=true2026.04 | 50.7 | |
| Qwen2.5 OmniReasoning=false, Model Size=3B, Zero-shot=true2026.04 | 50.2 | |
| AVATARReasoning=true, Model Size=7B, Zero-shot=true2026.04 | 49.1 | |
| AV-ReasonerReasoning=true, Model Size=7B, Zero-shot=true2026.04 | 48.3 | |
| Omni-R1Reasoning=true, Model Size=7B, Zero-shot=true2026.04 | 46.9 | |
| EchoInkReasoning=true, Model Size=7B, Zero-shot=true2026.04 | 46.5 | |
| AVATARReasoning=true, Model Size=3B, Zero-shot=true2026.04 | 45.8 | |
| Ola-7BReasoning=false, Model Size=7B, Zero-shot=true2026.04 | 45.3 | |
| HumanOmniReasoning=true, Model Size=7B, Zero-shot=true2026.04 | 44.9 | |
| v-SALMONN-o1Reasoning=true, Model Size=7B, Zero-shot=true2026.04 | 40.5 | |
| V-RTSReasoning=true, Model Size=7B, Zero-shot=true2026.04 | 35.3 | |
| Kimi-VL-ThinkingReasoning=true, Model Size=Teacher, Zero-shot=true2026.04 | 33.5 | |
| AF3 (think)Reasoning=true, Model Size=Teacher, Zero-shot=true2026.04 | 28.9 |