Skip to content

Benchmark

LongVideoBench

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

Models measured
6
Scores recorded
7
Sources
2
IQ weight
2.00

Score distribution

75.4%
Mean
79.8%
Median
8.7%
Std dev
74.8%
10th percentile
80.6%
90th percentile
54.7%
Min
54.7%80.6%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.00 for chat models, measured over 6 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Doubao Seed 2.1 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim80.6%
2Doubao Seed 2.1 Turbo· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim80.6%
3Kimi K2.5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim79.8%
4Kimi K2.5setting not named by the source· Reported by Moonshot AI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.79.8%
5Doubao Seed 1.8· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim77.4%
6Seed 2.0 Mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim74.8%
7Qwen2.5-VL 7B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim54.7%

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.