multimodal benchmark
TreeBench
TreeBench evaluates visual grounded reasoning, requiring models to localize and reason about fine-grained visual details.
Updated Aug 11, 2026
Models2
Model coverage2
MetricScore
EvidenceB
TreeBench Ranking
Higher score ranks better on this benchmark.
| 01 | BY | 71.1% | 100.0% | 2 | C | |
| 02 | BY | 71.1% | 0.0% | 2 | C |
TreeBench Score Distribution
A closer view of the leading scores on this benchmark.
TreeBench Highlights
The leading models and scores on this benchmark.
What is TreeBench?
What TreeBench measures and how its scores work.
TreeBench evaluates visual grounded reasoning, requiring models to localize and reason about fine-grained visual details.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- TreeBench
- Modality
- multimodal
- Primary category
- multimodal
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- treebench|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about TreeBench.
Which model scores highest on TreeBench?
Seed 2.1 Pro is currently ranked first with 71.1%.
What does TreeBench measure?
TreeBench evaluates visual grounded reasoning, requiring models to localize and reason about fine-grained visual details.
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.