API platform ↗
StepFun’s multimodal foundation model: API access for text, image and video understanding, coding and tool use.
TOOLORA / Local reading edition
Step 5 Preview
StepFun’s multimodal foundation model: API access for text, image and video understanding, coding and tool use.
Overview
Step 5 Preview is StepFun’s multimodal foundation model for coding and professional knowledge work. Developers submit text, images or video through the official API, using up to 1M tokens of context for documents, charts or code tasks with text output. It supports reasoning settings, structured output and tool calls; the integrating application supplies search, execution and local-file access. This entry opens the official API platform.
- Company country / region
- China · Registered office
Access options
Usage guide & costs
- Access & setup
Personal accounts on the China platform require identity verification before creating an API key. Use model ID step-5-preview with the OpenAI-compatible request format.
Official details ↗- Billing model
Hosted API billing follows input and output token use. Step Plan is a separate subscription route with its own entitlements and access conditions.
Official details ↗- Before you start
Current hosted docs specify 1M context and up to 64k output. Self-deployment uses StepFun Model License 1.0, including commercial-license conditions for certain MaaS uses, rather than unconditional MIT/Apache terms.
Official details ↗ · Official details ↗
Official sources and references
- Official websitehttps://platform.stepfun.com/
- Documentationhttps://platform.stepfun.com/docs/zh/guides/models/step-5-preview
- Documentationhttps://platform.stepfun.com/docs/zh/quickstart/overview
- Documentationhttps://github.com/stepfun-ai/Step-5-Preview
- Pricing sourcehttps://platform.stepfun.com/docs/zh/guides/pricing/details
Descriptions summarize public information; consult official websites for features, pricing and regional availability.