Deploy DeepSeek‑V3 LLM on Alibaba Cloud with One‑Click Model Gallery

DeepSeek‑V3 Model Overview

DeepSeek‑V3 is a Mixture‑of‑Experts (MoE) large language model released by DeepSeek, with a total of 671 billion parameters and 370 billion active parameters per token. It adopts Multi‑head Latent Attention (MLA) and the DeepSeekMoE architecture, introduces a loss‑free load‑balancing strategy, and uses multi‑token prediction as a training objective to improve performance. The model was pretrained on 14.8 trillion high‑quality tokens and subsequently refined through supervised fine‑tuning (SFT) and reinforcement learning.

PAI‑Model Gallery Overview

The Model Gallery is a component of Alibaba Cloud’s AI platform PAI that aggregates high‑quality pretrained models from domestic and international open‑source communities, covering LLM, AIGC, computer vision, and NLP domains. By adapting these models to PAI, users can train, deploy, and infer with zero‑code effort, streamlining the entire AI development workflow.

Access the PAI‑Model Gallery at https://pai.console.aliyun.com/#/quick-start/models.

One‑Click Deployment of DeepSeek‑V3

Log in to the PAI console.

Select the appropriate region in the top‑left corner.

In the left navigation, choose a workspace, then open the desired workspace.

Navigate to Quick Start > Model Gallery .

In the model list, click the DeepSeek‑V3 card to open its detail page.

Click Deploy in the upper‑right corner, configure the inference service name and resource specifications, and deploy to the PAI‑EAS inference platform.

After deployment, go to PAI‑Model Gallery > Task Management > Deployment Tasks , click the service name, then select View WEB Application to interact via the ChatLLM WebUI.

The deployed service also supports API inference; refer to the tutorial “Deploy LLM with EAS in 5 minutes” for usage.

For further model updates and support, users can join the PAI‑Model Gallery community via DingTalk (group 79680024618).