V2X-QA - Ego Prediction: leaderboard

V2X-QA (2026) asks multimodal models four-option driving questions about real vehicle-mounted and roadside camera views of the same scenes, answered with one option letter on the expert-verified test split. This board is the vehicle-side task on short-horizon prediction of salient agents or ego motion. Accuracy in percent; higher is better.

Metric: Accuracy (%) on the vehicle-side Ego Prediction task (VS4); four-option multiple-choice questions on the expert-verified test split, answered as a single option letter under the paper's unified prompt; higher is better. Source: arxiv.org. Saturation forecast: Not forecast. 10 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Qwen 3.5 Plus58.9#250
2Qwen3-VL-Flash55.8#734
3Gemini 2.5 Pro46.8#280
4GPT-5 Mini (2025-08-07)44.5#352
5Intern-S144.2#627
6Qwen 3 VL 30B A3B Instruct43.4#903
7Intern-S1-mini43.4#1186
8GPT-5.243#190
9Qwen 2.5 VL 72B Instruct41.9#849
10Gemini 3 Flash (Preview)38.9#156

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=v2x-qa-ego-prediction · How It Works · Data refreshed daily, snapshot 2026-11-22.