multimodal benchmark
MMSIBench
MMSIBench is a multimodal spatial-intelligence benchmark evaluating spatial reasoning over images.
Updated Aug 11, 2026
Models2
Model coverage2
MetricScore
EvidenceB
MMSIBench Ranking
Higher score ranks better on this benchmark.
| 1 | BY | 35.9% | 100.0% | 2 | C | |
| 2 | BY | 31.4% | 0.0% | 2 | C |
MMSIBench Score Distribution
A closer view of the leading scores on this benchmark.
MMSIBench Highlights
The leading models and scores on this benchmark.
What is MMSIBench?
What MMSIBench measures and how its scores work.
MMSIBench is a multimodal spatial-intelligence benchmark evaluating spatial reasoning over images.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MMSIBench
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- mmsibench|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about MMSIBench.
Which model scores highest on MMSIBench?
Seed 2.1 Pro is currently ranked first with 35.9%.
What does MMSIBench measure?
MMSIBench is a multimodal spatial-intelligence benchmark evaluating spatial reasoning over images.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
2 model results are currently shown.
Does this benchmark affect the overall score?
No. This benchmark is shown for reference but does not contribute to the overall score.