Alibaba Cloud Goes Full-Stack Agentic, Shifts Growth Engine
Alibaba Cloud has unveiled a comprehensive, full-stack Agentic transformation spanning chips, cloud infrastructure, models, and inference, signaling a strategic pivot to reposition the primary cloud consumer from humans to autonomous agents. The announcement introduces three major pillars—the Zhenwu M890 AI chip, the Qwen3.7-Max flagship model, and the Qianwen Cloud website—all designed to re-architect the cloud consumption model around token-based metering and agent-native workflows.
At the chip layer, the Zhenwu M890 delivers three times the performance of its predecessor, forming part of a fully self-developed data center chip matrix that underscores Alibaba Cloud's deepening vertical integration. At the model layer, the Qwen3.7-Max has demonstrated its autonomy by writing chip code with a tenfold performance improvement, illustrating the model's advanced reasoning and code generation capabilities. These components are not isolated upgrades but rather synchronized building blocks in a unified stack engineered to serve agent-driven workloads at scale.
The platform layer marks an equally significant shift. Alibaba Cloud's Bailian platform has been fully opened, welcoming partners including Moonshot, Zhipu, and others into a growing ecosystem. This openness positions the platform as an aggregation layer for diverse AI capabilities, supporting multi-model orchestration and agent deployment. On the business model side, MaaS (Model as a Service) revenue is poised to become Alibaba Cloud's largest product line, with AI-related revenue share now exceeding 30%. This formal transition to token-based metering—charging by consumption rather than by instance or VM—reflects a deeper economic restructuring designed to align with agent-native usage patterns, where billing is fluid, granular, and tightly coupled to model inference costs.
