Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Spatial Reasoning research benchmarks – Page 5 · SOTA2 Research

Spatial Reasoning

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

63.7Accuracy

13

Mar 12, 2026

RoboInter-VQA Spatial

10.9Grasp Pose Error

13

Feb 26, 2026

85.41GPT-grid Score

12

Jul 1, 2026

Real-world Robot Workstation Unseen Viewpoints

78.2Accuracy (45° Viewpoint)

12

Jun 25, 2026

GR-ER 1.5 Thinking

79.3Accuracy

12

May 5, 2026

RoboSpatial (val)

66.7Accuracy (RoboSpatial val)

12

Apr 10, 2026

69.3Test Accuracy

12

Apr 7, 2026

SAT-Dynamic-13B

87.7Val Metric Score

12

Apr 7, 2026

82.2Subject Accuracy

12

Apr 7, 2026

70.6Count Score

12

Apr 7, 2026

SPAR-Bench tiny

72.32Medium Difficulty Score

12

Mar 10, 2026

SIBench single-image

58.86Accuracy

12

Feb 26, 2026

COCO 2017 (val)

LLaVA (Group 1 Baseline)

83Alignment Accuracy

12

Feb 26, 2026

70.81Accuracy

12

Mar 24, 2026

MindCube Tiny Around

75.6Accuracy

11

Jun 2, 2026

VLM4D-Real Overall

82Accuracy (%)

11

Jun 2, 2026

VLM4D-Real Exo-centric

76Accuracy

11

Jun 2, 2026

VLM4D-Real Ego-centric

88Accuracy

11

Jun 2, 2026

RefSpatial (val)

48.4Accuracy

11

Apr 10, 2026

MMSI-Bench 79 (MV)

3DThinker-S1+S2

43.3Accuracy

11

Mar 9, 2026

ViewSpatial-Bench SI, MV 34

3DThinker-S1+S2

68.6Accuracy

11

Mar 9, 2026

SPAR-Bench SI, MV 91

3DThinker-S1+S2

63.3Accuracy

11

Mar 9, 2026

3DThinker-S1+S2

81.1Accuracy

11

Mar 9, 2026

SPBench SI, MV 35

SpatialLadder-3B

70.6Accuracy

11

Mar 9, 2026

VSI-Bench MV 76

3DThinker-S1+S2

63.7Accuracy

11

Mar 9, 2026

Page 5 of 9

Popular datasets

BLINK CV-Bench MindCube VSI-Bench SAT

Follow for updates

Join SOTA2 Contact