Deploy DeepSeek LLM on Huawei Cloud: Step‑by‑Step Guide with Ollama, ChatBox & CodeArts

This article explains how to deploy the DeepSeek distilled model on a Huawei Developer Space cloud host, covering Ollama installation, DeepSeek deployment, ChatBox integration, and CodeArts IDE setup.

1. Install Ollama

Ollama is an open‑source tool that lets you run and manage large language models locally. Open a terminal on the cloud desktop (right‑click → “Open Terminal Here”) and run: curl -fsSL https://ollama.com/install.sh | sh After the script finishes, Ollama is ready.

2. Deploy DeepSeek

Use Ollama to pull and run the DeepSeek‑r1:1.5b model with the following command: ollama run deepseek-r1:1.5b When the model starts, you can interact with it directly in the terminal.

3. Install ChatBox

Download ChatBox from https://chatboxai.app/zh using the Firefox browser on the cloud host, then give the AppImage execution permission: sudo chmod +x Chatbox-1.9.8-x86_64.AppImage Install the required library: sudo apt install libfuse2 Run ChatBox: ./Chatbox-1.9.8-x86_64.AppImage In ChatBox, choose “Use My Own API Key/Local Model”, select the Ollama API, and pick the deepseek-r1:1.5b model from the list, then save.

4. Integrate DeepSeek into CodeArts IDE

Install the Continue extension in CodeArts IDE (Extensions → search “Continue”). After installation, open the extension, click “Add Chat model”, choose Provider = Ollama, Model = Autodetect, and connect. The extension will automatically detect the previously deployed DeepSeek model.

Once connected, you can use DeepSeek inside CodeArts for code completion, generation, explanation, error checking, and knowledge queries, greatly enhancing development productivity.