Alibaba Cloud has bolstered its artificial intelligence (AI) capabilities for international customers, unveiling a suite of new models, tools and infrastructure upgrades.
Announced at the company’s Spring Launch 2025 online event, the new offerings focus on providing scalable AI offerings, particularly in the rapidly evolving field of generative AI (GenAI) and large language models (LLMs).
“We are launching a series of platform-as-a-service [PaaS] and AI capability updates to meet the growing demand for digital transformation from across the globe,” said Selina Yuan, president of international business at Alibaba Cloud. “These upgrades allow us to deliver even more secure and high-performance services that empower businesses to scale and innovate in an AI-driven world. As cloud and AI become essential for global growth, we are committed to enhancing our core product offerings to address our customers’ evolving needs.”
Key to the announcement is the expanded access to Alibaba Cloud’s foundational models, including the latest iterations of its proprietary LLM series, Qwen. Available through the company’s availability zones in Singapore, these models include the large-scale mixture of experts (MoE) model Qwen-Max, the reasoning model QwQ-Plus, the visual reasoning model QVQ-Max, and the multimodal model Qwen2.5-Omni-7b.
QwQ-Plus is designed for deep analytical thinking, tackling complex question-and-answer tasks and expert-level mathematical problems with algorithm-driven precision. QVQ-Max, on the other hand, focuses on visual reasoning, addressing complex multimodal problems with enhanced accuracy and extended reasoning capabilities. It supports both visual input and chain-of-thought output.
Supporting these advanced models is a significantly upgraded Platform for AI (PAI), Alibaba Cloud’s machine learning platform. PAI-Elastic Algorithm Service (EAS) now boasts distributed inference capabilities with a multi-node architecture to handle the increasing demands of super-large models, particularly those leveraging MoE structures and ultra-long-text processing.
Furthermore, PAI-EAS introduces a prefill-decode disaggregation function, boosting performance and reducing costs. Alibaba Cloud claimed that this function leads to a 92% increase in concurrency and a 91% increase in tokens per second when deployed with the Qwen2.5-72B model.
PAI-Model Gallery has also received a substantial upgrade, now offering nearly 300 open-source models. This includes the full range of Alibaba Cloud’s open-source Qwen and Wan series, all accessible through a no-code deployment and management interface. The gallery offers diverse deployment methods and underlying computing resources, along with new features such as model evaluation for performance insights and model distillation to reduce deployment costs.
To enable efficient data management in the AI era, Alibaba Cloud has also integrated its native AI inference capabilities, powered by Qwen, into its flagship cloud-native relational database PolarDB.
The integration will eliminate the need for data shifting typically required for inference workflows, leading to reduced latency and improved efficiency and data security. This is particularly suited for text-centric workloads, including conversational retrieval-augmented generation (RAG) agent development, text embedding generation and semantic similarity search.
Beyond infrastructure and platform enhancements, Alibaba Cloud also introduced a suite of software-as-a-service AI products, including AI Doc, an intelligent document processing tool that leverages LLMs to parse various document types, extract information, and generate tailored reports.
The latest updates come on the heels of Alibaba’s February 2025 announcement of a $53bn investment over three years to advance its cloud computing and AI infrastructure. Alibaba Cloud currently operates a global infrastructure network spanning 87 availability zones across 29 regions.
According to Synergy Research Group, the global public cloud market reached $330bn in 2024, up $60bn from 2023. The research firm noted that generative AI has driven half of the market growth over the past two years, through a combination of new platform and services, GPU as a service, and enhancements to a wide range of other cloud services. During the fourth quarter of 2024, Alibaba Cloud’s market share stood at 4%, behind Amazon, Microsoft and Google.