Skip to content

Benchmark

Wild Bench

Scores are percentages.

Models measured
8
Scores recorded
8
Sources
1
IQ weight
1.01

Score distribution

58.6%
Mean
65.3%
Median
9.5%
Std dev
48.5%
10th percentile
68.5%
90th percentile
42.4%
Min
42.4%68.5%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.01 for chat models, measured over 8 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Ministral 3 14B Instruct2512· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim68.5%
2Mistral Large 3· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim68.5%
3Ministral 3 8b Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim66.8%
4Mistral Small 3.2 24B Instruct2506· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim65.3%
5Ministral 3 3b Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim56.8%
6Mistral Small 24B Instruct2501· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim52.2%
7Jamba 1.5 Large· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim48.5%
8Jamba 1.5 Mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim42.4%

Data credits: Data by LLM Stats