Embodied3DBench - Grasp Point (2D): leaderboard

Embodied3DBench (2026) Grasp Point (2D) task on simulated robotic tabletop scenes: grasp points in the image plane, predicted zero-shot by a vision-language model. Normalized 0-100 score derived from box IoU or point distance; higher is better.

Metric: Grasp Point (2D) task (grasp points in the image plane), 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: Not forecast. 13 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1Qwen 3 VL 4B Instruct84.9#1332
2Qwen 2.5 VL 72B Instruct70.8#841
3GPT-566.2#178
4Claude Sonnet 450.7#410
5Qwen 2.5 VL 7B Instruct49.1#1902
6GPT-4o46.2#751
7Gemini 2.5 Pro35.4#276
8Gemini 2.5 Flash34.4#463
9InternVL3.5-8B33.9#1604
10InternVL3-38B31.5#939
11Doubao-Seed-1.6Thinking28.3
12Doubao-1.5-thinking-vision-pro27.6#344
13Seed-1.6Non-reasoning27#469

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

Interactive version: theaggregate.ai/benchmark?slug=embodied3dbench-grasp-point-2d · How It Works · Data refreshed daily, snapshot 2026-11-21.