StepFun's Step 5 Preview early tests: Pareto frontier on cost vs. capability

StepFun_ai · x · 2026-09-21

omarsar0 shared early hands-on impressions of StepFun's new Step 5 Preview model, testing it mainly as a coding agent.

His verdict: it sits on the Pareto frontier for cost vs. capability, is comparable to GLM 5.3, Kimi K3, and others, and comes at a very competitive price — "worth trying in your favorite coding agent." StepFun's official account thanked him. One of the first third-party test reports on this model.

Related event: StepFun launches Step 5 Preview, touting cost-performance frontier(3 posts)→

Original post →

More from coding & agent

coding & agent channel →