E3VS-Bench: leaderboard

E3VS-Bench (2026) tests vision-language models as embodied agents that move a free 5-DoF camera through photorealistic 3D Gaussian Splatting scenes to reach a viewpoint that answers a question; a fixed GPT-5.1 answerer and judge score the outcome 5 (correct) or 1 (incorrect) over 377 test episodes. Mean score on a 1-5 scale; higher is better.

Metric: Average over the six question categories of the judge score on the E3VS-Bench test split (377 episodes, 21 3D Gaussian Splatting scenes): the model drives a 5-DoF camera agent (at most 25 steps) to a viewpoint from which the question can be answered; GPT-5.1 then answers from the agent's final view and a GPT-5.1 judge scores the answer 5 (correct) or 1 (incorrect); mean judge score on a 1-5 scale; higher is better. Source: arxiv.org. Saturation forecast: Around March 2027. 8 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Gemini 3 FlashNon-reasoning2.79#175
2Gemini 3 Pro2.75#145
3Gemini 2.5 Pro2.54#276
4Qwen 3 VL 8B2.46#917
5GPT-5.1Non-reasoning2.42#240
6Qwen 3 VL 30B A3B2.32#784
7Step3 VL 10B2.24#1080
8InternVL3.5-8B2.21#1604

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=e3vs-bench · How It Works · Data refreshed daily, snapshot 2026-11-21.