Flow - 4x faster than typing
Mini toolsStep 3.7 Flash
By StepFun
Model family: Step
Step 3.7 Flash is a sparse Mixture of Experts model with 198B total parameters and 11B active parameters, offering a 256K token context window. It natively processes images and video without requiring a separate vision model, making it usable directly in agent frameworks. The model supports three selectable reasoning intensities, low, medium, and high, letting users trade off speed against depth for tasks ranging from simple Q&A to complex math, planning, and code analysis. It is designed for high throughput, low latency inference suited to real time agentic workflows, with reliable tool calling and multi step task decomposition. Typical use cases include code generation, whiteboard to plan conversion, chart to data extraction, and mobile screen based agent operation.
New
Multimodal
Overview
StepFun's flagship multimodal reasoning model, a 198B total parameter sparse MoE with 11B active parameters. Natively understands images and video, supports adjustable reasoning effort, and is optimized for agent and coding workflows with reliable tool calling.
About StepFun
StepFun AI is a creative AI assistant platform that offers chat, search, coding help, and multimodal generation (images, audio, video), plus research models like Step3 and NextStep-1.
Industry:
Artificial Intelligence
Company Size:
400
Location:
Shanghai, Xuhui District, CN
Website:
stepfun.ai
Last updated: July 14, 2026
