Embodied3DBench - Grounding (3D): leaderboard

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

Metric: Grounding (3D) task (metric 3D pointing and 3D bounding boxes), 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
1GPT-531.7#178
2Gemini 2.5 Pro30.7#276
3Seed-1.6Non-reasoning30#469
4Gemini 2.5 Flash26.3#463
5Doubao-Seed-1.6Thinking25.7
6Claude Sonnet 423.8#410
7Doubao-1.5-thinking-vision-pro23.5#344
8Qwen 3 VL 4B Instruct21#1332
9GPT-4o17.1#751
10Qwen 2.5 VL 72B Instruct12.9#841
11InternVL3-38B11.7#939
12InternVL3.5-8B10.4#1604
13Qwen 2.5 VL 7B Instruct5.3#1902

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

Interactive version: theaggregate.ai/benchmark?slug=embodied3dbench-grounding-3d · How It Works · Data refreshed daily, snapshot 2026-11-21.