StepFun
StepFun Open Platform
StepFun Open Platform is the developer entry for StepFun. The English docs cover OpenAI-compatible chat completion, model listing, files, token counting, tool calls, image generation and editing, TTS, ASR, voice cloning, pricing and agreements. The current homepage now highlights Step 3.7 Flash as the flagship multimodal reasoning model, while the broader text lineup still includes step-3.5-flash, step-2 and step-1 models.
Quick answers
At a glance
- Overview
- StepFun's English developer platform for Step 3.7 Flash, Step text, reasoning, audio, image, file and tool-call APIs.
- Best fit
- Developers comparing Chinese model APIs for text, reasoning, tool calling, multimodal generation and OpenAI-compatible migration.
- Trust
- 4/4 sources verified, recently checked · 2026-05-29
- Coverage
- 100/100
Editorial verdict
Best for
Developers comparing Chinese model APIs for text, reasoning, tool calling, multimodal generation and OpenAI-compatible migration.
Avoid if
Avoid relying on it blindly when procurement requires card billing, regional SLA or enterprise data terms without account-level confirmation.
Why it matters
The English platform makes StepFun more actionable for overseas developers than a company-only profile.
Pricing
Reasoning models start at $0.10 input cache miss / $0.02 cache hit / $0.30 output per 1M tokens; image editing is $0.003 per image
Payment
Account balance, Free credit first, Paid balance, WeChat Pay, Stripe for overseas users via Step Plan
Commercial use
Commercial use should follow the StepFun Open Platform terms, model-specific pricing and Step Plan terms where applicable.
Privacy
The English docs publish privacy, terms and data-processing agreement pages; review them before sensitive workloads.
Use-case fit
OpenAI-compatible API migration
StrongUse the documented chat-completions path when migrating model calls from OpenAI-style SDKs.
Tool-calling apps
StrongThe docs include tool-call support for applications that need external systems or actions.
Multimodal API comparison
MediumCompare StepFun against MiniMax, Qwen Cloud and Z.ai for audio, image and model-platform breadth.
Flagship multimodal reasoning
StrongUse Step 3.7 Flash when the platform needs one model for images, video, tool calls and long-context reasoning.
Global user checklist
Model list, flagship model positioning, plan benefits and pricing are changing quickly; verify current docs and account console before production.
Pros
- - Homepage now highlights Step 3.7 Flash as the flagship model
- - English docs cover text, reasoning, audio, image and tool-call APIs
- - OpenAI-compatible migration path is documented
- - Pricing and rate-limit pages are public
Cons
- - Some global signup and billing details still need account-level testing
- - Model availability differs between direct API and Step Plan subscription
Decision paths
It is the current homepage hero and the best single-model entry for multimodal reasoning.
Step Plan uses a subscription quota model for agent and coding tools.
MiniMax has mature international docs across text, speech, video, image and music.
Sources
official · en · verified 2026-05-29
Confirms the English platform entry point and Step 3.7 Flash homepage highlight.
docs · en · verified 2026-05-29
Lists API reference, model guides, pricing and Step Plan integration docs.
docs · en · verified 2026-05-29
Documents the flagship multimodal reasoning model, pricing, effort levels and framework support.
pricing · en · verified 2026-05-29
Documents reasoning, speech pricing and tiered rate limits.