evals.report
BenchmarksLabsCompareRun guides
StepFunStepFun Step series

Step 3.7 Flash

StepFun · StepFun Step series. Released May 29, 2026.

5 results

Benchmark results 5

Compare this model
BenchmarkCategoryScoreMetricStatusDate
SWE-bench VerifiedCoding76.5%% resolvedVerifiedMay 29, 2026Details
SWE-bench ProCoding56.3%% resolvedVerifiedMay 29, 2026Details
Terminal-Bench 2.1Agents59.6%task successVerifiedMay 29, 2026Details
Artificial Analysis Intelligence IndexReasoning42.6IndexUnverifiedMay 29, 2026Details
GDPvalAgents1298EloOfficialMay 29, 2026Details