LLM Benchmarks — Scores, Methodology & Live Rankings

What each benchmark measures, how models score, and the current rankings

Reset

Most tested benchmarks

Ranked by how many models have tracked scores — open a benchmark for its full live ranking.

ERQA 21 models tested · latest result 26.08.2026 72.3% CountBench 13 models tested · latest result 22.04.2026 97.8% EmbSpatialBench 12 models tested · latest result 22.04.2026 84.6% RefSpatialBench 12 models tested · latest result 22.04.2026 70% RefCOCO (avg) 11 models tested · latest result 22.04.2026 92.5% LingoQA 8 models tested · latest result 01.02.2026 82% ODInW13 6 models tested · latest result 16.04.2026 50.8% Hypersim 6 models tested · latest result 01.02.2026 13.1% Nuscene 6 models tested · latest result 01.02.2026 15.4% SUNRGBD 5 models tested · latest result 01.02.2026 36.2%

ODInW13

Spatial Intelligence

ODInW13 - object detection in the wild across 13 domains.

Active View →

EmbSpatialBench

Spatial Intelligence

EmbSpatialBench - embodied spatial intelligence benchmark.

Active View →

RefSpatialBench

Spatial Intelligence

RefSpatialBench - referring spatial reasoning benchmark.

Active View →

RefCOCO (avg)

Spatial Intelligence

RefCOCO (avg) - referring expression comprehension benchmark, average score.

Active View →

ERQA

Spatial Intelligence

ERQA - embodied/spatial reasoning question answering benchmark.

Active View →

CountBench

Spatial Intelligence

CountBench - visual object counting benchmark.

Active View →

Hypersim

Spatial Intelligence

Hypersim - 3D spatial understanding benchmark using synthetic indoor scenes.

Active View →

Nuscene

Spatial Intelligence

Nuscene - 3D spatial understanding benchmark using autonomous driving scenes.

Active View →

LingoQA

Spatial Intelligence

LingoQA - spatial and linguistic reasoning benchmark for visual grounding.

Active View →

SUNRGBD

Spatial Intelligence

SUNRGBD - 3D spatial understanding benchmark using RGB-D indoor scenes.

Active View →

RoboSpatial

Spatial Intelligence

RoboSpatial as evaluated on the LFM3-Omni-2B card: RoboSpatial-Home with the official RoboSpatial-Eval rule.

Active View →

Point-Bench

Spatial Intelligence

Point-Bench as evaluated on the LFM3-VL-2B card (vision benchmark, with-thinking setting): Visual grounding category, with-thinking column. Pointing benchmar...

Active View →

MindCube (tinybench)

Spatial Intelligence

MindCube (tinybench) as evaluated on the LFM3-VL-2B card (vision benchmark, with-thinking setting): Embodied reasoning category, with-thinking column. MindCu...

Active View →

SAT (val + test)

Spatial Intelligence

SAT (val + test) as evaluated on the LFM3-VL-2B card (vision benchmark, with-thinking setting): Embodied reasoning category, with-thinking column. SAT spatia...

Active View →

RoboSpatial-Home

Spatial Intelligence

RoboSpatial-Home as evaluated on the LFM3-VL-2B card (vision benchmark, with-thinking setting): Embodied reasoning category, with-thinking column. RoboSpatia...

Active View →

Where2Place

Spatial Intelligence

Where2Place as evaluated on the LFM3-VL-2B card (vision benchmark, with-thinking setting): Embodied reasoning category, with-thinking column. Relative spatia...

Active View →

OpenEQA (episodic memory, video)

Spatial Intelligence

OpenEQA (episodic memory, video) as evaluated on the LFM3-VL-2B card (vision benchmark, with-thinking setting): Embodied reasoning category, with-thinking co...

Active View →

VSI-Bench

Spatial Intelligence

VSI-Bench as evaluated on the LFM3-VL-2B card (vision benchmark, with-thinking setting): Embodied reasoning category, with-thinking column. Visual-Spatial In...

Active View →

CountQA

Spatial Intelligence

CountQA, counting QA over images (counting objects in pictures as typed numeric decisions). Values reported as percentage. Source referenced on the LiquidAI/...

Active View →