Integrating DeepSeek Large Model with Spring AI: A Step‑by‑Step Guide

DeepSeek is a domestically developed large language model released by DeepSeek Co., offering two main model families: the V series for dialogue (model name deepseek-chat ) and the R series for reasoning (model name deepseek-reasoner ).

Spring AI can integrate DeepSeek by reusing the existing OpenAI client. To get started you need a DeepSeek API key, set the base URL to api.deepseek.com , and choose the desired model via the spring.ai.openai.chat.model property.

Preparation steps:

Create an API key on the DeepSeek portal and configure it with the property spring.ai.openai.api-key .

Set the base URL using spring.ai.openai.base-url to api.deepseek.com .

Select the model by setting spring.ai.openai.chat.model=<model name> (e.g., deepseek-chat or deepseek-reasoner ).

Example integration:

1. Add the Maven dependency

<dependency>
    <groupId>org.springframework.ai</groupId>
    <artifactId>spring-ai-openai-spring-boot-starter</artifactId>
</dependency>

2. Configure Spring Boot

spring:
  ai:
    openai:
      api-key: sk-xxx   // replace with your own key
      base-url: https://api.deepseek.com
      chat:
        options:
          model: deepseek-chat

3. Simple chat controller (blocking and streaming)

package com.ivy.controller;

import org.springframework.ai.chat.messages.UserMessage;
import org.springframework.ai.chat.model.ChatResponse;
import org.springframework.ai.chat.prompt.Prompt;
import org.springframework.ai.openai.OpenAiChatModel;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;
import reactor.core.publisher.Flux;
import java.util.Map;

@RestController
public class ChatController {

    private final OpenAiChatModel chatModel;

    public ChatController(OpenAiChatModel chatModel) {
        this.chatModel = chatModel;
    }

    @GetMapping("/ai/generate")
    public Map
generate(@RequestParam(value = "message", defaultValue = "Tell me a joke") String message) {
        return Map.of("generation", this.chatModel.call(message));
    }

    @GetMapping("/ai/generateStream")
    public Flux
generateStream(@RequestParam(value = "message", defaultValue = "Tell me a joke") String message) {
        Prompt prompt = new Prompt(new UserMessage(message));
        return this.chatModel.stream(prompt);
    }
}

Because DeepSeek's public service may be limited by resources, you can also deploy the model locally for experimentation.

Conclusion : Integrating DeepSeek with Spring AI is straightforward, supporting both synchronous and streaming chat modes. The usage patterns for function calls, role definitions, and structured outputs are consistent with other Spring AI tutorials.

Source code repository: https://github.com/Fj-ivy/spring-ai-examples