DeepEyes: benchmark results
Provider: Other.
Aggregate Rating 1573 ± 24, rank #1362 of 10916 rated models, from 62 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| DiCoBench - Instance Commonality | 75.5 | Accuracy (%) on the instance-commonality task (re-identify o | 94.1 |
| DiCoBench - Entity Difference | 55.8 | Accuracy (%) on the entity-difference task (a small object a | 85.3 |
| VTC-Bench (Interface Tools) - Measurement | 43.81 | Pass rate (%) on measurement (105 questions on size, positio | 83.3 |
| DiCoBench - Reasoning Difference | 26.7 | Accuracy (%) on the reasoning-difference task (traces that i | 79.4 |
| DiCoBench - Category Commonality | 60.9 | Accuracy (%) on the category-commonality task (match objects | 70.6 |
| VTC-Bench (Code Tools) - Perceptual Restoration | 28 | Pass rate (%) on perceptual restoration (50 questions under | 63.9 |
| 360Bench (CMP) - Fine-Grained Perception | 39 | Fine-grained perception accuracy (%), mean of instance recog | 62.5 |
| VTC-Bench (Interface Tools) - Spatial Reasoning | 46.67 | Pass rate (%) on spatial reasoning under extreme conditions | 58.3 |
| DiCoBench - Attribute Difference | 31.8 | Accuracy (%) on the attribute-difference task (a subtle chan | 55.9 |
| DiCoBench - Spatial Commonality | 29.6 | Accuracy (%) on the spatial-commonality task (locate the sha | 53.1 |
| 360Bench (CMP) | 35.5 | Overall accuracy (%) over all seven subtasks on 360Bench's 1 | 50 |
| 360Bench (ERP) - Fine-Grained Perception | 39.3 | Fine-grained perception accuracy (%), mean of instance recog | 50 |
Interactive version: theaggregate.ai/model?slug=deepeyes · How It Works · Data refreshed daily, snapshot 2026-11-08.