Skip to content

Benchmark

Step 3.5 Search Agent · Average Tool Gain

Agentic benchmarks measure how well models carry out tasks that take several steps. Higher scores are better.

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

Score distribution

52
Mean
52
Median
0
Std dev
52
10th percentile
52
90th percentile
52
Min

Results

The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.

ModelScore
Step 3.5 FlashStep 3.5 technical report · Table 11 · 256K · mean across five search benchmarks· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.52

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.