Skip to content

Benchmark

Hypersim

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

Models measured
5
Scores recorded
5
Sources
2
IQ weight
1.42

Score distribution

2.6%
Mean
0.13%
Median
5%
Std dev
0.11%
10th percentile
12.5%
90th percentile
0.11%
Min
0.11%12.5%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.42 for chat models, measured over 5 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Qwen3.5 397B A17BQwen3.5 release/model card · published setting· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.12.5%
2Qwen3.5 35B A3B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim0.131%
3Qwen3.5 27B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim0.13%
4Qwen3.5 122B-A10B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim0.127%
5Qwen3 VL 235Bthinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.0.11%

Data credits: Data by LLM Stats · 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.