Audio Reasoning on MMAR
83.7Average AccuracyGemini-3.1-Pro
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Gemini-3.1-ProSize=–, Model Category=Closed-source models2026.06 | 83.7 | — | — | — | — | — | — | — | |
| Gemini-3-ProSize=–, Model Category=Closed-source models2026.06 | 81.73 | — | — | — | — | — | — | — | |
| VISASize=Agent2026.06 | 77.4 | 71.5 | 62.6 | 84 | 63.6 | 86.2 | 81.7 | 75 | |
| AudioToolAgentSource=Closed Source2025.10 | 77 | 73.33 | 63.59 | 82.99 | 63.64 | 83.03 | 81.71 | 79.17 | |
| Gemini 3 ProSource=Closed Source2025.10 | 76.4 | 72.73 | 61.65 | 84.35 | 72.73 | 81.19 | 78.05 | 83.33 | |
| Step Audio R1Size=32B2026.06 | 71.5 | 61.2 | 52.4 | 81.3 | 54.6 | 80.3 | 81.7 | 79.2 | |
| OmniJigsawvariant=CMM2026.04 | 71 | — | — | — | — | — | — | — | |
| EChO-AgentSize=–2026.06 | 71 | 63.64 | 53.4 | 81.29 | 81.82 | 77.06 | 74.39 | 75 | |
| AudioJigsaw2026.04 | 70.7 | — | — | — | — | — | — | — | |
| Qwen3-Omni ThinkingSize=30B2026.06 | 69.9 | 67.3 | 51 | 77.6 | 54.6 | 78.9 | 73.2 | 70.8 | |
| Qwen3-Omni Flash InstructSize=-2026.06 | 69.8 | 66.7 | 58.3 | 74.2 | 81.8 | 78.4 | 64.6 | 70.8 | |
| OmniJigsawvariant=SMS2026.04 | 69.5 | — | — | — | — | — | — | — | |
| SAR-LMSize=Agent2026.06 | 69.3 | 52.7 | 56.3 | 81 | — | — | — | — | |
| Step-Audio-R1Size=33B, Model Category=Open-source models2026.06 | 69.15 | — | — | — | — | — | — | — | |
| OmniJigsawvariant=JMI2026.04 | 69.1 | — | — | — | — | — | — | — | |
| Qwen-3-Omni-thinkingSize=30B2026.06 | 69 | 66.67 | 49.51 | 79.93 | 63.64 | 74.31 | 71.95 | 62.5 | |
| AudioToolAgentSize=Agent2026.06 | 68.8 | 61.8 | 51.9 | 77.6 | 72.7 | 76.6 | 72 | 70.8 | |
| Step-Audio-R1.1Size=33B, Model Category=Open-source models2026.06 | 68.75 | — | — | — | — | — | — | — | |
| Qwen-3-Omni-instructSize=30B2026.06 | 68.7 | 58.79 | 52.91 | 78.23 | 72.73 | 77.06 | 70.73 | 70.83 | |
| Qwen3-Omni-30B2026.04 | 68.5 | — | — | — | — | — | — | — | |
| Gemini 2.5 FlashSize=-2026.06 | 68.4 | 60 | 53.4 | 77.2 | 63.6 | 76.2 | 69.5 | 75 | |
| Qwen3-Omni Flash ThinkingSize=-2026.06 | 68.1 | 65.5 | 51.9 | 75.5 | 63.6 | 76.2 | 65.9 | 70.8 | |
| MOSS-Audio-8B-ThinkingSize=8B, Model Category=Open-source models2026.06 | 66.53 | — | — | — | — | — | — | — | |
| Qwen3-OmniReasoning=true, Model Size=30B, Zero-shot=true2026.04 | 66.4 | — | — | — | — | — | — | — | |
| Qwen3-Omni-30B-A3B-InstructSize=30B, Model Category=Open-source models2026.06 | 66.4 | — | — | — | — | — | — | — | |
| AudioToolAgent-OpenSource=Open Source2025.10 | 65.9 | 61.21 | 57.28 | 69.05 | 54.55 | 69.27 | 78.05 | 66.67 | |
| Gemini 2.0 FlashSource=Closed Source2025.10 | 65.6 | 61.21 | 50.97 | 72.11 | 81.82 | 72.48 | 65.85 | 70.83 | |
| Gemini 2.0 FlashZero-shot=true2026.06 | 65.6 | 61.2 | 51 | 72.1 | 81.8 | 72.5 | 65.9 | — | |
| Gemini 2.0 FlashSize=–2026.06 | 65.6 | — | 50.97 | 72.11 | 81.82 | 72.48 | 65.85 | 70.83 | |
| MOSS-Audio-8B-InstructSize=8B, Model Category=Open-source models2026.06 | 64.42 | — | — | — | — | — | — | — | |
| MOSS-Audio-4B-ThinkingSize=4B, Model Category=Open-source models2026.06 | 63.91 | — | — | — | — | — | — | — | |
| GPT-4o AudioSource=Closed Source2025.10 | 63.5 | 53.94 | 50.97 | 70.41 | 63.64 | 72.48 | 62.2 | 75 | |
| GPT-4o AudioSize=-2026.06 | 63.5 | 53.9 | 51 | 70.4 | 63.6 | 72.5 | 62.2 | 75 | |
| Omni-R1Source=Closed Source2025.10 | 63.4 | 67.3 | 51.5 | 64.3 | 45.5 | 70.2 | 64.6 | 70.8 | |
| MOSS-Audio-4B-InstructSize=4B, Model Category=Open-source models2026.06 | 62.54 | — | — | — | — | — | — | — | |
| HumanOmniV22026.04 | 61.8 | — | — | — | — | — | — | — | |
| MiMo-Audio-7BSize=7B, Model Category=Open-source models2026.06 | 61.7 | — | — | — | — | — | — | — | |
| AudioToolAgent-OpenSize=Agent2026.06 | 61.7 | 59.4 | 45.6 | 67.3 | 54.6 | 70.6 | 59.8 | 70.8 | |
| Kimi-AudioSize=7B, Model Category=Open-source models2026.06 | 60.82 | — | — | — | — | — | — | — | |
| AF3 (think)Reasoning=true, Model Size=Teacher, Zero-shot=true2026.04 | 60.1 | — | — | — | — | — | — | — | |
| GPT4o-AudioSize=–, Model Category=Closed-source models2026.06 | 59.78 | — | — | — | — | — | — | — | |
| Omni-R12026.04 | 59.7 | — | — | — | — | — | — | — | |
| AVRTReasoning=true, Model Size=7B, Zero-shot=true2026.04 | 59.1 | — | — | — | — | — | — | — | |
| AudioGenie-ReasonerSize=Agent2026.06 | 58.9 | 49.7 | 43.3 | 69.2 | 45.5 | 64.5 | 65.3 | 59.1 | |
| Audio Flamingo 3Size=8.3B, Setting=Base (reported)2026.06 | 58.5 | — | — | — | — | — | — | — | |
| Audio Flamingo 3Size=7B, Model Category=Open-source models2026.06 | 57.96 | — | — | — | — | — | — | — | |
| AVRTReasoning=true, Model Size=3B, Zero-shot=true2026.04 | 57.3 | — | — | — | — | — | — | — | |
| Qwen2.5-OmniSource=Open Source2025.10 | 56.7 | 58.79 | 40.78 | 59.86 | 54.55 | 61.93 | 67.07 | 58.33 | |
| Qwen2.5-OmniSize=7B, Model Category=Open-source models2026.06 | 56.7 | — | — | — | — | — | — | — | |
| Qwen2.5-OmniModel Size=7B, Zero-shot=true2026.06 | 56.7 | 58.8 | 40.8 | 59.9 | 54.6 | 61.9 | 67.1 | — | |
| Qwen-2.5-OmniSize=7B2026.06 | 56.7 | 58.79 | 40.78 | 59.86 | 54.55 | 61.93 | 67.07 | 58.33 | |
| Qwen2.5 OmniReasoning=false, Model Size=7B, Zero-shot=true2026.04 | 56.5 | — | — | — | — | — | — | — | |
| SpectCount (Audio Flamingo 3)Size=8.3B, Trained Parameters=26.2M, Setting=SpectCount2026.06 | 56.3 | — | — | — | — | — | — | — | |
| MiniCPM-o-4.5Size=9B, Model Category=Open-source models2026.06 | 55.75 | — | — | — | — | — | — | — | |
| Qwen-2.5-OmniSize=3B2026.06 | 53.8 | 53.94 | 46.12 | 53.74 | 36.36 | 60.09 | 57.32 | 58.33 | |
| Qwen2.5 OmniReasoning=false, Model Size=3B, Zero-shot=true2026.04 | 53.7 | — | — | — | — | — | — | — | |
| Audio Flamingo 3Size=8.3B, Setting=Base (reproduced)2026.06 | 52.9 | — | — | — | — | — | — | — | |
| v-SALMONN-o1Reasoning=true, Model Size=7B, Zero-shot=true2026.04 | 51.2 | — | — | — | — | — | — | — | |
| Audio Flamingo NextSize=8B, Model Category=Open-source models2026.06 | 51.01 | — | — | — | — | — | — | — | |
| GPT-4o mini AudioZero-shot=true2026.06 | 50.6 | 38.8 | 35.9 | 58.8 | 45.5 | 60.1 | 57.3 | — | |
| R1-AQASize=7B2026.06 | 47.6 | 55.8 | 37.4 | 49 | 9.1 | 50 | 50 | 50 | |
| Cap+DeepSeek-V3Model Size=671B, Zero-shot=true2026.06 | 47.6 | 42.4 | 40.8 | 56.1 | 18.2 | 50 | 45.1 | — | |
| TinyGiantALMModel Size=1.5B, Zero-shot=true2026.06 | 46.4 | 47.3 | 37.9 | 46.9 | 45.5 | 49.1 | 58.5 | — | |
| SpectCount (Qwen2-Audio-Instruct)Size=8.4B, Trained Parameters=25.9M, Setting=SpectCount2026.06 | 45.7 | — | — | — | — | — | — | — | |
| Baichuan-Omni-1.5Model Size=11B, Zero-shot=true2026.06 | 40.7 | 41.2 | 33 | 40.5 | 36.4 | 48.6 | 39 | — | |
| Baichuan-Omni-1.5Size=11B2026.06 | 40.7 | 41.21 | 33.01 | 40.48 | 36.36 | 48.62 | 39.02 | 41.67 | |
| Qwen2-Audio-InstructSize=8.4B, Setting=Base (reproduced)2026.06 | 40.1 | — | — | — | — | — | — | — | |
| Audio ReasonerSource=Open Source2025.10 | 36.8 | 43.64 | 33.5 | 32.99 | 45.45 | 42.66 | 31.71 | 25 | |
| Audio-ReasonerSize=7B2026.06 | 36.8 | 43.6 | 33.5 | 33 | 45.5 | 42.7 | 31.7 | 25 | |
| Audio-ReasonerModel Size=8.4B, Zero-shot=true2026.06 | 36.8 | 43.6 | 33.5 | 33 | 45.5 | 42.7 | 31.7 | — | |
| SALMONNSize=13B2026.06 | 33.2 | 30.3 | 31.1 | 34.7 | 9.1 | 34.9 | 35.4 | 41.7 | |
| SALMONNModel Size=13B, Zero-shot=true2026.06 | 33.2 | 30.3 | 31.1 | 34.7 | 9.1 | 34.9 | 35.4 | — | |
| Audio-CoTSize=7B2026.06 | 31.3 | 35.8 | 25.2 | 34 | 9.1 | 30.7 | 30.5 | 37.5 | |
| Audio-CoTModel Size=8.4B, Zero-shot=true2026.06 | 31.3 | 35.8 | 25.2 | 34 | 9.1 | 30.7 | 30.5 | — | |
| Qwen2-Audio-InstructSize=8.4B, Setting=Base (reported)2026.06 | 30 | — | — | — | — | — | — | — | |
| Qwen2-AudioModel Size=8.4B, Zero-shot=true2026.06 | 30 | 33.3 | 24.3 | 32.3 | 9.1 | 31.2 | 30.5 | — | |
| Random GuessSize=-2026.06 | 29.32 | 29.39 | 25.88 | 31.48 | 25 | 29.3 | 31.1 | 28.13 | |
| OpenOmniSize=8B2026.06 | 27 | 20.61 | 22.33 | 35.37 | 18.18 | 27.06 | 23.17 | 25 | |
| GAMAModel Size=7B, Zero-shot=true2026.06 | 26.5 | 29.1 | 24.3 | 27.9 | 27.3 | 24.8 | 28.1 | — | |
| AnyGPT-chatSize=8B2026.06 | 23.7 | 24.24 | 19.42 | 22.11 | 27.27 | 27.52 | 26.83 | 29.17 | |
| Flamingo 2Model Size=3B, Zero-shot=true2026.06 | 21.9 | 24.9 | 17.5 | 20.8 | 18.2 | 26.6 | 23.2 | — | |
| LTU-ASModel Size=7B, Zero-shot=true2026.06 | 19 | 20 | 14.1 | 19.1 | 9.1 | 20.6 | 28.1 | — |