Key configuration (click to copy)
Base URL
https://api.deepseek.com
Chat / Edit Model
deepseek-chat
Code Model
deepseek-coder
Setting up VS Code Continue with DeepSeek
Who it's for
- Developers who want to validate AI coding with a low-cost setup first
- Those more comfortable with VS Code than Cursor
- Anyone who wants the option to migrate to other compatible APIs later
Prerequisites
- VS Code installed
- The Continue extension installed
- A DeepSeek platform account with an API Key
Key configuration
- Base URL:
https://api.deepseek.com - Chat / Edit Model:
deepseek-chat - Code Model:
deepseek-coder
Setup steps
- Open the Continue configuration file.
- Add a new OpenAI-compatible provider.
- Enter the DeepSeek Base URL, API Key, and default model.
- Use a simple repository to verify code explanation, completion, and refactoring.
Troubleshooting
- If Continue cannot read the configuration, check the JSON syntax first.
- If model responses don't match expectations, make sure the chat model and the code model are configured separately.
- If costs rise too quickly, reduce autocomplete frequency and limit large-context Q&A.
Cost note
This is one of the best paths for low-budget experimentation today. Light-to-moderate usage typically stays within ¥0-80/month.
配置完成?Setup complete?
前往官方平台开通服务,开始使用 AI 编程。Head to the official site to activate the service and start AI coding.