Skip to content

Benchmark

StepFun Consulting & Recommendations

Scores are percentages. Higher is better.

Models measured
1
Scores recorded
5
Sources
1
IQ weight
No weight published

Score distribution

70.5%
Mean
72.4%
Median
4.4%
Std dev
62.1%
10th percentile
74.2%
90th percentile
62.1%
Min
62.1%74.2%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1Step 3.5 FlashStep 3.5 technical report · internal evaluations · instruction following· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.74.2%
2Step 3.5 FlashStep 3.5 technical report · internal evaluations · usefulness· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.73.3%
3Step 3.5 FlashStep 3.5 technical report · internal evaluations · tone· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.72.4%
4Step 3.5 FlashStep 3.5 technical report · internal evaluations · average· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.70.5%
5Step 3.5 FlashStep 3.5 technical report · internal evaluations · logic· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.62.1%

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.