StepFun

Step Plan is StepFun's subscription-based AI service for Agent and Coding scenarios. It offers standardized APIs with OpenAI and Anthropic compatibility, intelligent multi-model routing via step-router-v1, and out-of-the-box multimodal capabilities including text, speech, and image models. Designed for developers building intelligent workflows with high-frequency model calls.

Step Plan is a subscription-based AI service from StepFun for high-frequency developers, offering access to flagship models via standardized APIs for coding and agent scenarios with intelligent routing and multimodal capabilities.

Step 3.5 Flash is a fast-response language model optimized for Chinese language understanding and generation. It is designed for quick inference and efficient performance in conversational and text-based tasks.

Step 3.5 Flash is a fast and efficient language model from StepFun, optimized for low-latency responses. It is part of the Flash series, designed to balance speed with strong reasoning capabilities for general-purpose tasks.

Step3 VL 10B is a multimodal vision-language model developed by StepFun. With 10 billion parameters, it is designed to understand and process both visual and textual information for various tasks.

Automatic speech recognition model for streaming and near-realtime transcription.

Text-to-speech model with zero-shot voice cloning and natural-language control.

Lightweight generative editing model for text-to-image and image editing with fast response.

Intelligent routing model for automatic switching between deepseek-v4-pro and step-3.5-flash based on task complexity.