en.Wedoany.com Reported - Step-1 Star has released the Step Edge on-device model family, comprising four products: the Step Edge base model, Step Edge Audio, Step Edge GUI, and Step Edge Gen, targeting terminal scenarios such as smartphones and automobiles.

This model suite aims to extend artificial intelligence agents from the cloud to terminal devices, enabling faster local responses and stronger privacy protection, while handling complex tasks through cloud-device collaboration. The three core features of Step Edge include: ultra-low local latency, supporting on-device local toolcall execution as low as 0.1 seconds, reducing reliance on cloud links for simple, high-frequency tasks; full-modal privacy protection, processing multimodal information such as text, vision, and speech locally, ensuring sensitive data does not leave the device, suitable for privacy-sensitive scenarios like smartphones and automobiles; native cloud-device collaboration, where simple tasks and weak or no network environments are handled quickly on the device, while complex reasoning and long-chain tasks are delegated to the cloud, achieving a balance of speed, capability, and cost.

This series of models can be paired with the Step Inference NPU engine, optimized for inference on terminal hardware, further reducing end-to-end latency across various input modalities such as text, vision, and speech.