Alibaba Cloud launches support for Qwen 3.5 and GLM-5 for coding
Alibaba Cloud announced a major update to its platform, integrating four leading open models: Qwen 3.5, GLM-5, MiniMax M2.5, and Kimi K2.5. These neural network
AI-processed from Jiqizhixin (机器之心); edited by Hamidun News
Alibaba Cloud is turning into a showcase for Chinese AI: the platform has simultaneously integrated four leading open-source language models — Qwen 3.5, GLM-5, MiniMax M2.5, and Kimi K2.5. All four systems share one common feature: they were created specifically for programming tasks and demonstrate results that closely approach the best proprietary solutions on the market. For developers working with Alibaba's cloud infrastructure, this means access to an entire arsenal of tools without the need to switch between multiple providers.
The context of this announcement is important. Over the past year and a half, Chinese technology companies have been systematically releasing models that previously remained in the shadow of their Western counterparts. The Qwen series from Alibaba itself, GLM from Tsinghua University and Zhipu AI, the work of MiniMax and Moonshot AI teams — all of this is not scattered experiments, but a consistently built ecosystem.
Now Alibaba Cloud consolidates these developments under one roof, effectively claiming a role as the central hub for open-source AI not only in Asia, but on the global market. The timing is no accident: Western corporate users are increasingly seeking alternatives to expensive APIs from OpenAI and Anthropic, and regulatory uncertainty around closed models is pushing companies toward open solutions with predictable licensing policies.
Each of the four models occupies its own niche in the new offering. Qwen 3.5 is the flagship of Alibaba's own lineup, enhanced for code generation and debugging tasks across a multitude of programming languages.
GLM-5 from Zhipu AI bets on complex logical inference and work with multi-step instructions, which is critical for automating software development. MiniMax M2.5 brings extended long-context capabilities to the bundle — the ability to keep large codebases in memory as a whole without losing coherence of reasoning.
Kimi K2.5 from Moonshot AI fills the niche of agentic scenarios: the model can independently break down a task into subtasks, plan steps, and execute them sequentially, making it a promising foundation for autonomous coding agents. Together, the four systems cover virtually the entire spectrum of tasks that a development team faces: from writing individual functions to complex refactoring of large projects.
The consequences of this move for the industry go beyond simply expanding the catalog. First, Alibaba Cloud creates direct competition to services like Amazon Bedrock and Azure AI Studio, which aggregate models from different manufacturers under a single API. The difference is that the Chinese platform focuses exclusively on open-source models, reducing customer dependency on the vendor and providing the option to deploy the same systems locally if needed.
Second, the concentration of top Chinese models in one place simplifies their testing and comparison for international corporations that are considering these developments but are not ready to build integrations with each provider separately. Third, the very fact of such consolidation signals the maturity of the Chinese AI segment: the market has moved from the stage of publication race to the stage of commercial deployment.
The practical benefit for developers is already evident. Instead of juggling accounts across Zhipu AI, MiniMax, and Moonshot, a single Alibaba Cloud account with a unified billing, monitoring, and scaling system is sufficient. This reduces operational burden and allows focus on the product rather than infrastructure administration.
The main question now facing the market is how sustainable this advantage will prove to be. The open-source landscape changes rapidly: models considered cutting-edge today risk ending up in the middle of benchmark tables in six months. Alibaba Cloud is betting not on specific checkpoints, but on a platform effect — the more developers get used to working with its infrastructure, the harder it will be for competitors to pull them to their side. If the platform continues to integrate new models as rapidly as they are released, it has every chance of cementing its status as the de facto standard for working with Chinese open-source systems on a global market scale.
Want to stop reading about AI and start using it?
AI News is a curated feed of AI/tech news. Hamidun Academy teaches you to use AI systematically in your work.