Integrate DeepSeek with Spring AI: Step‑by‑Step Spring Boot Guide

Spring AI Overview

Spring AI is an open‑source framework from the Spring ecosystem that simplifies AI integration for Java developers through abstracted, modular APIs, supporting major AI services such as OpenAI, DeepSeek, Google, and Ollama.

Unified abstraction API for various AI providers.

Core modules for model interaction, vector handling, retrieval‑augmented generation (RAG), and function calling.

Low‑code integration via Spring Boot Starter.

Structured output mapping model responses to Java objects.

Flux‑based streaming responses for real‑time chat.

Obtaining an API Key

Use Alibaba Cloud Bailei’s DeepSeek service. After locating the DeepSeek model at

https://bailian.console.aliyun.com

, click “立即体验” and then the key icon to generate your API key.

Usage

Add the Spring AI dependency to your Spring Boot project:

<code>&lt;dependency&gt;
    &lt;groupId&gt;org.springframework.ai&lt;/groupId&gt;
    &lt;artifactId&gt;spring-ai-openai-spring-boot-starter&lt;/artifactId&gt;
    &lt;version&gt;1.0.0-M6&lt;/version&gt;
&lt;/dependency&gt;</code>

Configure the AI service in

application.yml

:

<code>spring:
  ai:
    openai:
      # API key for the AI service
      api-key: &lt;YOUR_API_KEY&gt;
      # Base URL for Alibaba Cloud Bailei
      base-url: https://dashscope.aliyuncs.com/compatible-mode
      chat:
        options:
          # Model selection (e.g., deepseek-r1 or deepseek-v3)
          model: deepseek-r1
          # Controls randomness of generation (lower = more deterministic)
          temperature: 0.8</code>

Create a controller to expose two endpoints – one for direct answers and another for streaming responses:

<code>/**
 * @author macrozheng
 * @description Controller for DeepSeek integration
 */
@RestController
public class DeepSeekController {

    private final OpenAiChatModel chatModel;

    @Autowired
    public DeepSeekController(OpenAiChatModel chatModel) {
        this.chatModel = chatModel;
    }

    /** Return a direct answer */
    @GetMapping("/ai/chat")
    public Map chat(@RequestParam(value = "message") String message) {
        return Map.of("generation", this.chatModel.call(message));
    }

    /** Return a streaming answer */
    @GetMapping(value = "/ai/chatFlux", produces = MediaType.TEXT_EVENT_STREAM_VALUE + "; charset=UTF-8")
    public Flux<ChatResponse> chatFlux(@RequestParam(value = "message") String message) {
        Prompt prompt = new Prompt(new UserMessage(message));
        return this.chatModel.stream(prompt);
    }
}
</code>

Testing

Start the application and use Postman to call the endpoints:

Direct answer:

http://localhost:8080/ai/chat?message=YourQuestion

Streaming answer:

http://localhost:8080/ai/chatFlux?message=YourQuestion

The complete response text can be retrieved from the

result.output.text

property.

Conclusion

Integrating DeepSeek with Spring AI is straightforward. For faster responses, prefer the streaming endpoint and concatenate the streamed text.

Source Code

https://github.com/macrozheng/spring-examples/tree/master/spring-deepseek

References

Spring AI documentation: https://docs.spring.io/spring-ai/reference/api/chat/deepseek-chat.html

Alibaba Cloud Bailei DeepSeek integration guide: https://help.aliyun.com/zh/model-studio/developer-reference/deepseek