Skip to content

Benchmark

BLINK - Relative_Depth

Vision benchmarks measure how well models understand images. Scores are percentages.

Models measured
4
Scores recorded
4
Sources
1
IQ weight
No weight published

Score distribution

57.7%
Mean
58.9%
Median
5.2%
Std dev
50%
10th percentile
64.5%
90th percentile
50%
Min
50%64.5%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1LLaVA 1.6 34BRelative_Depth· EvalAI data64.5%
2Qwen-VL MaxRelative_Depth· EvalAI data58.9%
3Claude 3 OpusRelative_Depth· EvalAI data57.3%
4Gemini ProRelative_Depth· EvalAI data50%