Step 3.7 Flash
By stepfun
stepfun/step-3.7-flashStepFun’s flagship multimodal reasoning model. Powered by a 198B-parameter / 11B-activation sparse MoE architecture, with native support for image and video understanding.
ReleasedMay 28, 2026
Pricing
Pricing (Input)$0.20/M
Pricing (Output)$1.15/M
CachingYes
Limits
Context Length256K
Max Output Tokens256K
Input Modalities
TextSupported
ImageSupported
PDFNot supported
AudioNot supported
Output Modalities
TextSupported
ImageNot supported
AudioNot supported
Features
ReasoningSupported
ToolsSupported
TemperatureSupported