Artificial Intelligence is undeniably transforming smart devices, but many of today’s products have yet to move beyond treating AI as an add-on. A new generation of software is taking a different approach by embedding AI into the operating system itself. This emerging AI-native OS model aims to make interactions more natural, personalized, and proactive.
It is against this backdrop that Shanghai Droi Technology has introduced DroiClaw, an AI-native operating system developed for next-generation smart terminals. Rather than being merely an AI assistant or chatbot, DroiClaw is designed with AI at the foundation of its architecture.
Whereas traditional operating systems revolve around applications, with AI typically added as a separate assistant or productivity feature, an AI-native operating system integrates AI into the core of its framework. This influences how users interact with devices at a fundamental level, from how models are selected to the way tasks are processed.
DroiClaw uses multimodal large-model and AI technologies to support interactions through voice, text, images, documents, video, and natural-language commands. It also enables personalized AI avatars with customizable voices, personalities, memories, and skills. By embedding intelligence throughout the operating system, DroiClaw is designed to deliver a seamless experience across communication, work, travel, and everyday productivity.
One of DroiClaw’s key goals is to build an app-free agentic OS ecosystem, in which users do not need to switch between multiple applications. To that end, the operating system is designed to understand a user’s intent and coordinate tasks across services through AI agents.
Users can describe what they want in natural language while AI agents break requests into subtasks, call the necessary system functions, and complete authorized operations with fewer manual steps. While the vision is not to do away with apps overnight, this approach allows many routine interactions to become intent-driven rather than app-driven.
DroiClaw includes a skill library to support this strategy, wherein users can install and create AI skills for specific workflows. Features like AI-assisted writing, translation, meeting transcription, file analysis, and intelligent task automation exemplify the many ways AI agents can simplify routine activities without requiring users to navigate multiple interfaces.
DroiClaw combines lightweight on-device AI models with cloud-based large models to support different usage scenarios.
Under normal conditions, users primarily access cloud-based large models, which provide more advanced capabilities such as multimodal interactions, AIGC, complex reasoning, content generation, and skill execution.
On-device models mainly serve as a fallback when network access is unavailable or cloud service credits have been exhausted. In these situations, the system can automatically switch to a local model to handle basic question-and-answer tasks.
The OS also has the flexibility to support multiple AI models rather than locking users into a single provider. This open approach allows partners to tailor AI capabilities for industry-specific applications.
Transparency and user control become ever more essential as AI agents become capable of carrying out increasingly complex tasks. For that reason, DroiClaw aims to establish a secure, controllable, and observable AI agent ecosystem. The platform emphasizes user authorization for sensitive operations, permission management, intelligent task scheduling, and layered data protection.
Users gain greater visibility into how AI agents perform automated actions, and privacy-sensitive information can be processed locally where appropriate. These capabilities are intended to help build trust as AI evolves toward actively completing tasks on users’ behalf.
Beyond technological innovation, DroiClaw also reflects a broader shift toward AI democratization. The first smartphones preloaded with DroiClaw—launched with Coolpad and Philips in June 2026—are priced at around RMB 1,099, bringing advanced AI capabilities to more affordable smart devices.
The OS is also built as an open ecosystem, supporting multiple hardware platforms and service providers. While discussions around projects like OpenClaw indicate growing interest in open AI-agent technologies, DroiClaw represents its own vision for open, hardware-neutral AI-native platforms.
According to Shanghai Droi Technology, the company was founded in 2008 and has extensive experience in developing operating systems and mobile ecosystem services. Its FreemeOS platform has been installed on more than 200 million devices across dozens of countries and regions, with support from more than 1,000 hardware and software ecosystem partners.
DroiClaw is the next step, building on that experience to establish a vision for the next generation of intelligent computing; one that combines an AI-native OS architecture, an app-free agentic OS ecosystem, hybrid on-device and cloud AI, and a secure, controllable, and observable AI agent ecosystem.
As smart terminals evolve beyond the traditional app-based operating system, the platform is designed to make advanced AI more accessible through an open ecosystem and more affordable devices.