StepFun API Pricing

StepFun is one of the smaller Chinese labs, and its Flash-class models compete directly with the cheapest tiers from far larger providers.

Model Input / 1M Output / 1M Cached input / 1M Batch Notes
Step 3.5 Flash $0.10 $0.30 N/A N/A Low-cost tier from StepFun, competitive with the cheapest Gemini and Llama options.

StepFun has one model tracked here, Step 3.5 Flash, at $0.10 in / $0.30 out per 1M tokens. That ranks #7 of 60 models across 17 providers on a blended 3:1 rate.

One tier, priced to compete

Step 3.5 Flash is listed at $0.10 per million input tokens and $0.30 per million output. That is level with Gemini 2.5 Flash-Lite on input and a quarter cheaper on output, and it undercuts Llama 4 Scout on hosted providers. For a smaller lab, matching Google's floor price is the point: at this end of the market, price is most of the pitch.

Only Step 3.5 Flash is tracked here. StepFun ships more models than this, but the rest could not be confirmed with both an input and an output rate, and this page would be less useful with a guess on it.

What a model at this price is for

At ten cents per million input tokens, the arithmetic changes: a workload processing a hundred million input tokens a month costs about ten dollars in input. At that point the API bill is no longer the thing worth optimising, and the interesting questions become latency, reliability, and whether the model is accurate enough on your task to avoid a second pass.

Models in this band are built for classification, routing, extraction, and short completions rather than for reasoning. Reach for one when the task is well-specified and high-volume; reach for a flagship when a wrong answer is expensive to catch downstream.

Every StepFun model on this site

Model Blended / 1M Rank of 60 Full pricing
Step 3.5 Flash $0.150 #7 Step 3.5 Flash pricing page

StepFun pricing FAQ

How much does the StepFun API cost?

Step 3.5 Flash is $0.10 per million input tokens and $0.30 per million output, which places it among the cheapest models tracked on this site.

How does StepFun compare to Gemini Flash-Lite?

Step 3.5 Flash matches Gemini 2.5 Flash-Lite exactly on input at $0.10 per million, and comes in cheaper on output at $0.30 against Gemini's $0.40. Gemini's batch discount closes that gap for latency-tolerant work. Both are competing for the same high-volume, low-complexity workloads.

Related pages

Rates last checked on 2026-08-16, sourced from StepFun Platform Pricing. Standard public list prices only. See the methodology page for how figures are verified, and report anything that looks out of date.