豆包 AI 手机的崛起正挑战既有 App 生态体系
Counterpoint Research·2026-01-12 02:45

Core Insights - The article discusses the launch of the Nubia M153, the first native AI smartphone developed by ByteDance and ZTE, which was released as a "technology preview" on December 1, 2025, and sold out quickly, indicating strong market interest [4] - According to Counterpoint's "AI 360" report, the global smartphone shipment share supporting GenAI is expected to exceed 30% in 2025, up from 20% in 2024, and reach 57% by 2029, highlighting the growing importance of AI in high-end smartphones [5] - The article emphasizes the competitive landscape between smartphone brands and internet giants, focusing on user engagement, service entry control, and high-value data acquisition [8][10] Market Trends - The Nubia M153 is equipped with a Snapdragon 8 Elite platform, and future iterations are expected to enhance NPU performance to 80-100 TOPS, raising hardware requirements for smartphones [5] - AI functionalities are increasing the demand for RAM and ROM, as well as high-bandwidth memory, while also raising power consumption by approximately 8%-12%, posing challenges for battery technology [5] AI Capabilities - The core advantage of the Doubao AI smartphone lies in its ability to understand user intent and execute tasks across multiple apps, although it still faces challenges in complex, multi-app scenarios [6][8] - The high-privilege simulation operations of Doubao AI are causing friction with existing app ecosystems, triggering security mechanisms in major applications like WeChat and Alipay [6][8] Strategic Partnerships - The collaboration between ByteDance and ZTE mirrors the division of labor seen in other tech partnerships, with Doubao providing AI capabilities and Nubia handling hardware design [10] - The article suggests that smartphone brands with established ecosystems may not adopt the same division of labor, instead exploring open and collaborative app frameworks [10] Future Outlook - The article anticipates that smartphone brands and leading apps will gradually reach a consensus on standardized ecological interfaces, allowing system-level AI to act as a unified scheduling hub [10] - The control of high-value data and data security will remain central issues in ongoing negotiations between smartphone manufacturers and app developers [10]