Skip to content

Benchmark

$OneMillion-Bench · Expert Score

Higher scores are better.

Models measured
5
Scores recorded
5
Sources
1
IQ weight
0.44

Score distribution

49.7
Mean
52.5
Median
5.5
Std dev
41.8
10th percentile
55.9
90th percentile
41.8
Min
41.855.9
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.44 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
1Claude Fable 5Qwen表:expert score · Gemini 3.1 Pro Preview judge· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.55.9
2GPT-5.6 SolQwen表:expert score · Gemini 3.1 Pro Preview judge· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.53.8
3Qwen3.8 MaxQwen表:expert score · Gemini 3.1 Pro Preview judge· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.52.5
4Qwen3.7 MaxQwen表:expert score · Gemini 3.1 Pro Preview judge· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.44.4
5Claude Opus 4.8Qwen表:expert score · Gemini 3.1 Pro Preview judge· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.41.8

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.