Skip to content

Benchmark

Lexam Hard

Scores are percentages. Higher is better.

Models measured
10
Scores recorded
10
Sources
1
IQ weight
0.00

Score distribution

36.4%
Mean
37.5%
Median
3.2%
Std dev
31.9%
10th percentile
38.9%
90th percentile
29.5%
Min
29.5%40.8%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.00 for chat models, measured over 10 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Inkling FP4· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim40.8%
2DeepSeek V4 Pro· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim38.9%
3DeepSeek V4 Flash· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim38.2%
4GLM 5.2· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim37.5%
5GPT OSS 120B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim37.5%
6MiniMax M3· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim37.1%
7Nemotron 3 Ultra· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim36.7%
8Kimi K2.6· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim36.2%
9Mistral Medium 3.5 128b· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim31.9%
10Kimi K3· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim29.5%

Data credits: Hugging Face official benchmark leaderboards; licence varies by benchmark