Drive-P2D - Key Object (Object-1): leaderboard
Metric: Accuracy (%; identify the object in the scene that most influences the driving decision (single choice); zero-shot, all scenarios, mean of two runs). Source: arxiv.org. Saturation forecast: Estimated already saturated. 11 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 2.5 VL 72B Instruct | 68.06 |
| 2 | GPT-4.1 | 67.48 |
| 3 | Qwen 2.5 VL 7B Instruct | 65.58 |
| 4 | Gemma 3 27B (IT) | 64.46 |
| 5 | Gemini 2.5 Pro | 62.9 |
| 6 | Phi-4 Multimodal Instruct | 60.23 |
| 7 | Llama 3.2 11B Instruct | 58.67 |
| 8 | Llama 3.2 90B Vision Instruct | 51.39 |
| 9 | Gemma 3 4B (IT) | 49.86 |
Interactive version: theaggregate.ai/benchmark?slug=drive-p2d-key-object-object-1 · How It Works · Data refreshed daily, snapshot 2026-09-26.