Alibaba Launches DeepSeek AI Models on Cloud Service, Following Tech Giants
Alibaba Group Holding's cloud-computing services unit has introduced DeepSeek's artificial intelligence (AI) models on its platform, joining other major tech companies in offering the Chinese start-up's open-source systems to their clients. Alibaba Cloud stated that users can now access the entire process from training to deployment to inference with zero coding. The platform aims to simplify model development, providing developers and enterprise users with a faster, more efficient, and convenient AI development and application experience.

Alibaba Cloud users can access DeepSeek's AI models through its PAI Model Gallery, which includes the start-up's advanced AI models, DeepSeek-V3 and DeepSeek-R1. These models are known for being developed at a fraction of the cost and computing power typically required by major AI tech companies to build large language models (LLMs). The gallery also offers distilled versions of these models, such as DeepSeek-R1-Distill-Qwen-7B.
Large language models (LLMs) power generative AI services like OpenAI's ChatGPT. Open-source access allows third-party developers to modify or share the design of a software program, enhancing its capabilities. Distillation, a method of training smaller models to mimic larger ones while reducing computational costs, is common among companies aiming to scale down model sizes while maintaining performance.
Alibaba Cloud's recent move to offer DeepSeek's models, including the Qwen 2.5-Max model, reflects a trend among major tech companies to support the start-up's models for the benefit of their customers. Huawei Technologies' cloud-computing unit collaborated with AI infrastructure start-up SiliconFlow to make DeepSeek's V3 and R1 models available on its Ascend platform. Tencent Holdings and Nvidia have also integrated DeepSeek's models into their cloud-computing platforms.
Cloud computing technology allows enterprises to manage and distribute software and digital resources over the internet as an on-demand service. Tencent Holdings recently announced support for DeepSeek's R1 reasoning model on its cloud-computing platform, while Nvidia highlighted the capabilities of DeepSeek-R1 on its NIM microservice. Microsoft, Amazon, and other tech giants have also incorporated DeepSeek's models into their cloud services.
Despite the success of DeepSeek's cost-effective AI models, some experts question the significance of the breakthrough. Fudan University computer science professor Zheng Xiaoqing noted that the training expenditure for DeepSeek's V3 model excluded certain costs associated with prior research and experiments. Zheng highlighted that DeepSeek's success was attributed to engineering optimisation, which may not significantly impact chip purchases or shipments.
Alibaba Cloud introduces DeepSeek's AI models on its platform
Other tech giants like Huawei, Tencent, and Nvidia also support DeepSeek's models
DeepSeek's cost-effective AI models raise questions among experts about their significance
Source: SCMP