Embodied3DBench - Affordance (3D): leaderboard

Embodied3DBench (2026) Affordance (3D) task on simulated robotic tabletop scenes: functional contact point and interaction vector in metric 3D, predicted zero-shot by a vision-language model. Normalized 0-100 score derived from box IoU or point distance; higher is better.

Metric: Affordance (3D) task (functional contact point and interaction vector in metric 3D), normalized score (0-100): IoU of predicted boxes or the L2 error of predicted points (pixels in 2D, metres in 3D) mapped to 0-100, on the simulated robotic tabletop scenes of Embodied3DBench (21K QA pairs over 4.8K object instances), zero-shot; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 13 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Gemini 2.5 Pro54.8#280
2GPT-553.1#185
3Claude Sonnet 448.4#414
4Gemini 2.5 Flash43#466
5Seed-1.6Non-reasoning41.5#481
6Doubao-Seed-1.6Thinking39.5
7Doubao-1.5-thinking-vision-pro34.8#368
8GPT-4o30.7#754
9Qwen 3 VL 4B Instruct28.5#1348
10InternVL3.5-8B25.1#1618
11InternVL3-38B23.9#948
12Qwen 2.5 VL 72B Instruct22.2#849
13Qwen 2.5 VL 7B Instruct20#1919

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

Interactive version: theaggregate.ai/benchmark?slug=embodied3dbench-affordance-3d · How It Works · Data refreshed daily, snapshot 2026-11-22.