阶跃推出 Step Edge 端侧模型家族

Step Edge is a text and vision foundational model for edge-side scenarios such as mobile phones and vehicles, covering core edge-side capabilities such as image understanding, GUI grounding, tool invocation, and spatial/video understanding. The model can perform screen content understanding, interface element localization, text and visual information extraction, spatial reasoning, and multi-turn tool execution locally, supporting applications such as edge-side intelligent assistants, in-vehicle interaction, and local automation, and is suitable for low-latency, privacy-sensitive, and weak network environments.