Visual Question Answering on V2X-QA cooperative 1.0 (test)
92.1CO1 AccuracyV2X-MoE-8B
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| V2X-MoE-8BModel Type=Open-Source2026.04 | 92.1 | 93.2 | 83.4 | 75.9 | 86.2 | |
| Gemini-3-Flash-PreviewModel Type=Closed-Source2026.04 | 86 | 3.8 | 43.4 | 13.6 | 36.7 | |
| Intern-S1-MiniModel Type=Open-Source2026.04 | 64.5 | 29.1 | 43.4 | 5.3 | 35.6 | |
| Qwen-2.5-VL-72B-InstructModel Type=Open-Source2026.04 | 61.9 | 54.7 | 42.6 | 8.7 | 42 | |
| Intern-S1Model Type=Open-Source2026.04 | 57.4 | 70.2 | 43 | 3.4 | 43.5 | |
| Qwen-3.5-PlusModel Type=Closed-Source2026.04 | 37.7 | 30.9 | 64.5 | 2.8 | 34 | |
| GPT-5.2Model Type=Closed-Source2026.04 | 34 | 36.2 | 49.4 | 27.2 | 36.7 | |
| GPT-5-MiniModel Type=Closed-Source2026.04 | 26.4 | 9.8 | 47.2 | 11.3 | 23.7 | |
| Qwen3-VL-FlashModel Type=Closed-Source2026.04 | 22.3 | 47.9 | 44.9 | 6.5 | 30.4 | |
| Qwen-3-VL-30B-A3B-InstructModel Type=Open-Source2026.04 | 20.8 | 32.1 | 52.1 | 4.9 | 27.5 | |
| Gemini-2.5-ProModel Type=Closed-Source2026.04 | 15.8 | 59.2 | 45.3 | 3.4 | 30.9 |