← All models
StepFun: Step 3.7 Flash logo

Stepfun · Video

StepFun: Step 3.7 Flash

Step 3.7 Flash is StepFun's latest high-efficiency multimodal Mixture-of-Experts model. It pairs a 196B-parameter language backbone with a vision encoder for native image and video understanding, activating roughly 11B parameters...

ClosedIn: Text, Image, VideoOut: Text

Input / M

$0.20 / M

Output / M

$1.15 / M

Cache read / M

$0.04 / M

Context

262K

Released

May 28, 2026