Skip to content

Benchmark

StepFun Data Analysis Benchmark

Scores are percentages. Higher is better.

Models measured
1
Scores recorded
2
Sources
1
IQ weight
No weight published

Score distribution

39.6%
Mean
39.6%
Median
0.01%
Std dev
39.6%
10th percentile
39.6%
90th percentile
39.6%
Min
39.58%39.6%
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 · Claude Code · avg@3· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.39.6%
2Step 3.5 FlashStep 3.5 technical report · internal evaluations · Claude Code · avg@3 · prose precision· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.39.6%

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.