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

DeepSeek is a Chinese large‑language model offering two series: the V series for chat (model name deepseek-chat) and the R series for reasoning (model name deepseek-reasoner).

To integrate DeepSeek with Spring AI, first create an API key from the DeepSeek portal, then set the base URL property spring.ai.openai.base-url to api.deepseek.com and choose the desired model via spring.ai.openai.chat.model=<model name>.

Configuration can be added to application.yml (or properties) as follows:

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

A simple Spring Boot controller demonstrates two endpoints: /ai/generate for synchronous generation and /ai/generateStream for streaming responses. The controller injects OpenAiChatModel, builds a Prompt from a UserMessage, and returns either a map with the generated text or a Flux<ChatResponse> stream.

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

Because the public DeepSeek service may be subject to resource limits, the article also suggests deploying a DeepSeek model locally for uninterrupted learning and experimentation.

In summary, connecting DeepSeek to Spring AI is straightforward and enables both blocking and streaming chat capabilities with minimal configuration.