Skip to content

Benchmark

LOCA-Bench (256k)

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

Models measured
2
Scores recorded
4
Sources
2
IQ weight
No weight published

Score distribution

54.8%
Mean
60.3%
Median
5.5%
Std dev
49.3%
10th percentile
60.3%
90th percentile
49.3%
Min
49.3%60.3%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1MiniMax M3.1 Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim60.3%
2MiniMax M3.1 FlashMiniMax M3.1 Flash official-weights model-card run · reasoning effort, sampling, and run count not stated for this benchmark · official implementation · 256K· Reported by MiniMax· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.60.3%
3MiniMax M3· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim49.3%
4MiniMax M3official react mode · environment description 256K· Reported by MiniMax Research· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.49.3%

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links. · Data by LLM Stats