Step 3.7 Flash: benchmark results

StepFun's speed-oriented Step 3.7 Flash model. Provider: StepFun. Released 2026-05-29. Access: Open.

Aggregate Rating 1684 ± 10, rank #236 of 1656 rated models, from 189 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Tau2-Bench Telecom98.54Success Rate (%)99.2
AA TAU-2 Bench98.54Accuracy (%)99
ReferFlow-Bench - Shift - F-Acc45.1F-Acc (%)95.7
ReLE - Agents - BFCL-V372.3Accuracy (%)94.2
ReLE - Agents and Tool Use69.9Accuracy (%)93.6
AA Omniscience - Software Engineering (SWE) - Rust70Accuracy (%)92.2
CLAW-Eval80.5Score (%)92
AA Omniscience - Software Engineering (SWE) - Julia32Accuracy (%)90.1
AA Omniscience - Software Engineering (SWE) - Kotlin34Accuracy (%)88.1
AA Omniscience - Software Engineering (SWE) - HTML50Accuracy (%)87.7
ReferFlow-Bench - Refinement - T-Acc10.2T-Acc (%)87
ReferFlow-Bench - Shift - T-Acc22.3T-Acc (%)87

Interactive version: theaggregate.ai/model?slug=step-3-7-flash · How It Works · Data refreshed daily, snapshot 2026-10-20.