springboot接入deepseek深度求索 java

以下是在springboot中接入ai deepseek的过程。官网并没有java的示例。其实java的都可以用。

1. 创建 API key 网址:deepseek API keys

点击创建API key,把创建的key值复制下来,以后就不能再查看了,只能重新创建。

2. 封装询问deepseek的工具方法askDeepSeek()

添加key值和询问路径。API_KEY为你创建的key值。

复制代码
    private static final String API_URL = "https://api.deepseek.com/chat/completions";
    private static final String API_KEY = "11111111"; // 替换为你的 API Key

工具方法askDeepSeek():

传入的参数question就是要询问的问题。

复制代码
public String askDeepSeek(String question) throws IOException {
        CloseableHttpClient client = HttpClients.createDefault();

        // 创建 HTTP POST 请求
        HttpPost request = new HttpPost(API_URL);
        request.setHeader("Content-Type", "application/json");
        request.setHeader("Authorization", "Bearer " + API_KEY);

        // 动态构建请求体
        String requestBody = String.format(
                "{\"model\": \"deepseek-chat\", \"messages\": [{\"role\": \"user\", \"content\": \"%s\"}], \"stream\": false}",
                question
        );
        request.setEntity(new StringEntity(requestBody));

        // 发送请求并获取响应
        try (CloseableHttpResponse response = client.execute(request)) {
            // 返回响应内容
            return EntityUtils.toString(response.getEntity());
        }
    }

3. 调用该询问ai的方法

复制代码
String  question1= "今天是星期几。 " ;
String answer = askDeepSeek(question);
System.out.println("answer = " + answer);

4. 成功返回示例

content里的就是返回的回答。

复制代码
answer = {"id":"88dbce49-2841-448d-a74f-a2d3180c5672","object":"chat.completion","created":1734525002,"model":"deepseek-chat","choices":[{"index":0,"message":{"role":"assistant","content":"当然,我很高兴!谢谢你的关心。😊"},"logprobs":null,"finish_reason":"stop"}],"usage":{"prompt_tokens":12,"completion_tokens":11,"total_tokens":23,"prompt_cache_hit_tokens":0,"prompt_cache_miss_tokens":12},"system_fingerprint":"fp_1bcb2de9ac"}

不过我接入之后,他只能回答一些很简单的问题,有没有大佬会用啊~

相关推荐
invicinble2 小时前
对linux形成认识
linux·运维·服务器
技术路上的探险家2 小时前
8 卡 V100 服务器:基于 vLLM 的 Qwen 大模型高效部署实战
运维·服务器·语言模型
半桔3 小时前
【IO多路转接】高并发服务器实战:Reactor 框架与 Epoll 机制的封装与设计逻辑
linux·运维·服务器·c++·io
绵绵细雨中的乡音3 小时前
深入理解 ET 与 LT 模式及其在 Reactor 模型中的应用
服务器·网络·php
JH30733 小时前
SpringBoot 优雅处理金额格式化:拦截器+自定义注解方案
java·spring boot·spring
HABuo3 小时前
【linux文件系统】磁盘结构&文件系统详谈
linux·运维·服务器·c语言·c++·ubuntu·centos
Howrun7774 小时前
关于Linux服务器的协作问题
linux·运维·服务器
Coder_Boy_4 小时前
技术让开发更轻松的底层矛盾
java·大数据·数据库·人工智能·深度学习
invicinble5 小时前
对tomcat的提供的功能与底层拓扑结构与实现机制的理解
java·tomcat
较真的菜鸟5 小时前
使用ASM和agent监控属性变化
java