Alibaba Cloud has named Brazilian open-source specialist 4Linux as a local partner to bring its Qwen artificial intelligence models to businesses, as the Chinese group opens its first cloud region in Brazil and South America.
The partnership will combine Alibaba’s portfolio of Qwen models and cloud infrastructure with 4Linux’s deployment, integration and professional training, Alibaba said. Sao Paulo-based 4Linux, founded in 2001, describes its business as open-source IT consulting, support and technical training.
In a post on its LinkedIn company page, 4Linux said it would use open Qwen models and its local implementation experience to take generative AI to Brazilian companies. It also promoted a corporate large-language-model service that puts open models into production inside customers’ own environments, including deployment on Kubernetes and retrieval-augmented generation connected to internal data.
Open-weight models make their trained parameters available for download, allowing companies to run or adapt them without sending every request to the developer’s hosted service. That separates the reach of a model from sales of cloud capacity and managed AI products.
The arrangement could therefore give Alibaba a route to spread Qwen without requiring every workload to run on its own cloud. But the same flexibility leaves open whether wider use of the models will generate demand for the new Brazil region. Neither company has disclosed the partnership’s commercial terms, named a customer or said whether Alibaba Cloud is the required or preferred infrastructure.
Alibaba said the new Sao Paulo region features two data centres and offers computing, storage, containers, networking, security, databases, big data and cloud-native services. It also said it plans to introduce a suite of enterprise agentic-AI products.
The announcement did not list a locally available graphics-processing-unit instance, Model Studio service, Qwen endpoint or training cluster. Alibaba’s cloud documentation lists the Brazil region as sa-east-1, with two availability zones, but warns that regional support varies by product.
Alibaba also named Brazilian technology provider Insi as a partner for tailored cloud and AI services for large and medium-sized companies. The announcement did not say whether Insi would deploy Qwen, and neither local partner identified a live customer workload.
The Brazil opening follows Alibaba Cloud’s expansion in Mexico and other markets. Alibaba has committed to invest at least Rmb380bn, or $53bn, in AI and cloud infrastructure over three years, but has not disclosed how much of that sum is earmarked for Brazil.
The first named Qwen workloads will show whether the partnership primarily feeds Alibaba’s Brazil cloud or customer-controlled infrastructure. For now, the announcement establishes a local implementation route for the models; it does not establish a Brazilian Qwen business operating at scale.
