Lab release history
Last updated May 29, 2026
StepFun model releases
Chinese AI startup building Step multimodal and agentic foundation models. This page collects the lab's model releases, lifecycle events, source links, and model metadata in one crawlable record.
5 models
Step-3.7-Flash
AvailableStepFun's high-efficiency multimodal sparse-MoE successor to Step-3.5-Flash: a ~196B-total / ~11B-active vision-language model with native image and video understanding, a 256K context, and selectable reasoning tiers (high/medium/low). Tuned for coding agents and search workflows.
Step-3.5-Flash
AvailableStepFun's Apache-licensed sparse MoE model for fast agentic execution, coding, math, browsing, and tool-use workflows.
Step-2
AvailableSecond-generation StepFun foundation model line with larger-scale multimodal and reasoning ambitions.
Step-1V
AvailableStepFun's first major vision-language model, released after the Step-1 language model.
Step-1
AvailableStepFun's first public foundation model generation, introduced as a trillion-parameter Chinese model line.