学习链接

阿里AI-Spring Cloud Alibaba AI：快速搭建自己的通义千问

让你快速入门Spring Cloud Alibaba AI

SpringCloud + Spring AI Alibaba 整合阿里云百炼大模型

Spring AI Alibaba官方文档
spring-ai-alibaba 官方github
spring-ai-alibaba-examples - spring ai alibaba示例代码

spring-cloud-alibaba-examples/spring-cloud-ai-example（忽略这个）

Spring Cloud Alibaba官网
Spring Boot 单体应用升级 Spring Cloud 微服务最佳实践

阿里百炼

阿里百炼官网

创建api-key

查看api调用示例

jdk版本使用17（如果使用jdk8，okhttp版本和gson版本不兼容，要么这个类找不到，要么那个方法没有。更换了版本也总会有其它地方的问题。）

示例

pom.xml

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0"xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd"><modelVersion>4.0.0</modelVersion><!-- 用java8 总是 okhttp方法找不到，类找不到，有时这个版本能用，但另外的问题还存在，换了版本，方法又不存在。 --><parent><artifactId>spring-boot-dependencies</artifactId><groupId>org.springframework.boot</groupId><version>3.4.2</version></parent><groupId>org.example</groupId><artifactId>demo-alibailian</artifactId><version>1.0-SNAPSHOT</version><properties><maven.compiler.source>17</maven.compiler.source><maven.compiler.target>17</maven.compiler.target></properties><dependencies><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-web</artifactId></dependency><dependency><groupId>com.alibaba</groupId><artifactId>dashscope-sdk-java</artifactId><version>2.18.2</version><exclusions><!-- 排除slf4j-simple依赖 --><exclusion><groupId>org.slf4j</groupId><artifactId>slf4j-simple</artifactId></exclusion></exclusions></dependency></dependencies></project>

AQuickStart

package com.zzhua;// dashscope SDK的版本 >= 2.18.2import com.alibaba.dashscope.aigc.generation.Generation;
import com.alibaba.dashscope.aigc.generation.GenerationParam;
import com.alibaba.dashscope.aigc.generation.GenerationResult;
import com.alibaba.dashscope.common.Message;
import com.alibaba.dashscope.common.Role;
import com.zzhua.constant.Constants;import java.util.Arrays;/*
测试:
curl -X POST "https://dashscope.aliyuncs.com/api/v1/services/aigc/text-generation/generation" \
-H "Authorization: Bearer sk-~~~" \
-H "Content-Type: application/json" \
-d '{"model": "deepseek-r1","input":{"messages":[{"role": "user","content": "你是谁？"}]},"parameters": {"result_format": "message"}
}'
*/
public class AQuickStart {public static GenerationResult callWithMessage() throws Exception {Generation gen = new Generation();Message userMsg = Message.builder().role(Role.USER.getValue()).content("你是谁？").build();GenerationParam param = GenerationParam.builder()// 若没有配置环境变量，请用百炼API Key将下行替换为：.apiKey("sk-xxx").apiKey(Constants.API_KEY).model("deepseek-r1").prompt("你是谁？").messages(Arrays.asList(userMsg))// 不可以设置为"text".resultFormat(GenerationParam.ResultFormat.MESSAGE).build();return gen.call(param);}public static void main(String[] args) {try {GenerationResult result = callWithMessage();System.out.println("思考过程：");System.out.println(result.getOutput().getChoices().get(0).getMessage().getReasoningContent());System.out.println("回复内容：");System.out.println(result.getOutput().getChoices().get(0).getMessage().getContent());} catch (Exception e) {// 使用日志框架记录异常信息System.err.println("An error occurred while calling the generation service: " + e.getMessage());}System.exit(0);}
}

MultiChat

package com.zzhua;// dashscope SDK的版本 >= 2.18.2import java.util.Arrays;
import java.lang.System;import com.alibaba.dashscope.aigc.generation.Generation;
import com.alibaba.dashscope.aigc.generation.GenerationParam;
import com.alibaba.dashscope.aigc.generation.GenerationResult;
import com.alibaba.dashscope.common.Message;
import com.alibaba.dashscope.common.Role;
import com.alibaba.dashscope.exception.ApiException;
import com.alibaba.dashscope.exception.InputRequiredException;
import com.alibaba.dashscope.exception.NoApiKeyException;
import com.alibaba.dashscope.utils.JsonUtils;
import com.zzhua.constant.Constants;public class MultiChat {public static GenerationResult callWithMessage() throws ApiException, NoApiKeyException, InputRequiredException {Generation gen = new Generation();Message userMsg1 = Message.builder().role(Role.USER.getValue()).content("你好").build();Message AssistantMsg = Message.builder().role(Role.ASSISTANT.getValue()).content("你好！很高兴见到你，有什么我可以帮忙的吗？").build();Message UserMsg2 = Message.builder().role(Role.USER.getValue()).content("你是谁").build();GenerationParam param = GenerationParam.builder()// 若没有配置环境变量，请用百炼API Key将下行替换为：.apiKey("sk-xxx").apiKey(Constants.API_KEY).model("deepseek-r1").messages(Arrays.asList(userMsg1, AssistantMsg, UserMsg2))// 不可以设置为"text".resultFormat(GenerationParam.ResultFormat.MESSAGE).build();return gen.call(param);}public static void main(String[] args) {try {GenerationResult result = callWithMessage();System.out.println("思考过程：");System.out.println(result.getOutput().getChoices().get(0).getMessage().getReasoningContent());System.out.println("回复内容：");System.out.println(result.getOutput().getChoices().get(0).getMessage().getContent());} catch (ApiException | NoApiKeyException | InputRequiredException e) {// 使用日志框架记录异常信息System.err.println("An error occurred while calling the generation service: " + e.getMessage());}System.exit(0);}
}

StreamChat

package com.zzhua;// dashscope SDK的版本 >= 2.18.2import java.util.Arrays;import com.zzhua.constant.Constants;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import com.alibaba.dashscope.aigc.generation.Generation;
import com.alibaba.dashscope.aigc.generation.GenerationParam;
import com.alibaba.dashscope.aigc.generation.GenerationResult;
import com.alibaba.dashscope.common.Message;
import com.alibaba.dashscope.common.Role;
import com.alibaba.dashscope.exception.ApiException;
import com.alibaba.dashscope.exception.InputRequiredException;
import com.alibaba.dashscope.exception.NoApiKeyException;
import io.reactivex.Flowable;import java.lang.System;public class StreamChat {private static final Logger logger = LoggerFactory.getLogger(StreamChat.class);private static StringBuilder reasoningContent = new StringBuilder();private static StringBuilder finalContent = new StringBuilder();private static boolean isFirstPrint = true;private static void handleGenerationResult(GenerationResult message) {String reasoning = message.getOutput().getChoices().get(0).getMessage().getReasoningContent();String content = message.getOutput().getChoices().get(0).getMessage().getContent();if (!reasoning.isEmpty()) {reasoningContent.append(reasoning);if (isFirstPrint) {System.out.println("====================思考过程====================");isFirstPrint = false;}System.out.print(reasoning);}if (!content.isEmpty()) {finalContent.append(content);if (!isFirstPrint) {System.out.println("\n====================完整回复====================");isFirstPrint = true;}System.out.print(content);}}private static GenerationParam buildGenerationParam(Message userMsg) {return GenerationParam.builder()// 若没有配置环境变量，请用百炼API Key将下行替换为：.apiKey("sk-xxx").apiKey(Constants.API_KEY).model("deepseek-r1").messages(Arrays.asList(userMsg))// 不可以设置为"text".resultFormat(GenerationParam.ResultFormat.MESSAGE).incrementalOutput(true).build();}public static void streamCallWithMessage(Generation gen, Message userMsg)throws NoApiKeyException, ApiException, InputRequiredException {GenerationParam param = buildGenerationParam(userMsg);Flowable<GenerationResult> result = gen.streamCall(param);result.blockingForEach(message -> handleGenerationResult(message));}public static void main(String[] args) {try {Generation gen = new Generation();Message userMsg = Message.builder().role(Role.USER.getValue()).content("你是谁？").build();streamCallWithMessage(gen, userMsg);
//             打印最终结果
//            if (reasoningContent.length() > 0) {
//                System.out.println("\n====================完整回复====================");
//                System.out.println(finalContent.toString());
//            }} catch (ApiException | NoApiKeyException | InputRequiredException e) {logger.error("An exception occurred: {}", e.getMessage());}System.exit(0);}
}

Spring AI Alibaba

1、注意：因为 Spring AI Alibaba 基于 Spring Boot 3.x 开发，因此本地 JDK 版本要求为 17 及以上。

2、spring-ai-alibaba-examples：此仓库中包含许多 Example 来介绍 Spring AI Alibaba 从基础到高级的各种用法和 AI 项目的最佳实践。 更详细的介绍介绍请参阅每个子项目中的 README.md 和 Spring AI Alibaba 官网。

3、 阿里云百炼平台 获取 API-KEY

4、由于 spring-ai 相关依赖包还没有发布到中央仓库，如出现 spring-ai-core 等相关依赖解析问题，请在您项目的 pom.xml 依赖中加入指定的仓库配置。

简单示例

pom.xml

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0"xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd"><modelVersion>4.0.0</modelVersion><parent><artifactId>spring-boot-dependencies</artifactId><groupId>org.springframework.boot</groupId><version>3.4.2</version></parent><groupId>org.example</groupId><artifactId>demo-alibailian</artifactId><version>1.0-SNAPSHOT</version><properties><maven.compiler.source>17</maven.compiler.source><maven.compiler.target>17</maven.compiler.target></properties><dependencies><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-web</artifactId></dependency><!-- spring-ai-alibaba-starter 传递依赖了dashscope-sdk-java, 但是可能不是需要的版本（不是最新的版本） --><dependency><groupId>com.alibaba</groupId><artifactId>dashscope-sdk-java</artifactId><version>2.18.2</version><exclusions><!-- 排除slf4j-simple依赖 --><exclusion><groupId>org.slf4j</groupId><artifactId>slf4j-simple</artifactId></exclusion></exclusions></dependency><dependency><groupId>com.alibaba.cloud.ai</groupId><artifactId>spring-ai-alibaba-starter</artifactId><version>1.0.0-M5.1</version></dependency></dependencies><repositories><repository><id>spring-milestones</id><name>Spring Milestones</name><url>https://repo.spring.io/milestone</url><snapshots><enabled>false</enabled></snapshots></repository></repositories></project>

application.yml

spring:ai:dashscope:api-key: sk-~~~ # 阿里百炼的key

HelloworldController

@RestController
@RequestMapping("/helloworld")
public class HelloworldController {private static final String DEFAULT_PROMPT = "你是一个博学的智能聊天助手，请根据用户提问回答！";private final ChatClient dashScopeChatClient;// ChatClientAutoConfiguration中注册了scope为prototype的ChatClient.Builder,// 同时该ChatClient.Builder会依赖1个 ChatModel, 此ChatModel在DashScopeAutoConfiguration中定义了DashScopeChatModel这个bean,// 所以，当需要ChatClient.Builder时，就会触发多例bean的获取，从而会寻找ChatModel的实现bean，如果有不止1个ChatModel的bean定义了，那么就会报错public HelloworldController(ChatClient.Builder chatClientBuilder) {this.dashScopeChatClient = chatClientBuilder.defaultSystem(DEFAULT_PROMPT)// 实现 Chat Memory 的 Advisor// 在使用 Chat Memory 时，需要指定对话 ID，以便 Spring AI 处理上下文。.defaultAdvisors(new MessageChatMemoryAdvisor(new InMemoryChatMemory()))// 实现 Logger 的 Advisor.defaultAdvisors(new SimpleLoggerAdvisor())// 设置 ChatClient 中 ChatModel 的 Options 参数.defaultOptions(DashScopeChatOptions.builder().withTopP(0.7).build()).build();}// 也可以如下这样创建，与上面是1个意思/*public HelloworldController(ChatModel chatModel) {// 构造时，可以设置 ChatClient 的参数// {@link org.springframework.ai.chat.client.ChatClient};this.dashScopeChatClient = ChatClient.builder(chatModel)// 实现 Chat Memory 的 Advisor// 在使用 Chat Memory 时，需要指定对话 ID，以便 Spring AI 处理上下文。.defaultAdvisors(new MessageChatMemoryAdvisor(new InMemoryChatMemory()))// 实现 Logger 的 Advisor.defaultAdvisors(new SimpleLoggerAdvisor())// 设置 ChatClient 中 ChatModel 的 Options 参数.defaultOptions(DashScopeChatOptions.builder().withTopP(0.7).build()).build();}*/// http://localhost:8080/helloworld/simple/chat?query=你是谁/*** ChatClient 简单调用*/@GetMapping("/simple/chat")public String simpleChat(@RequestParam("query") String query) {// 1、DefaultChatClientRequestSpec 构造方法中添加的 CallAroundAdvisor 触发了请求的执行//    调用 chatModel.call(Prompt)，即 DashScopeChatModel#call(Prompt)// 2、然后调用 DashScopeApi#chatCompletionEntity(ChatCompletionRequest)//    可以看到 最终默认调用的模型是 qwen-plus//    （是因为 DashScopeAutoConfiguration 定义了 DashScopeChatModel 这个bean，//      而 DashScopeChatModel 这个bean又引用了 DashScopeChatProperties 这个bean，//      而 DashScopeChatProperties 中直接初始化的 DashScopeChatOptions 使用了 qwen-plus。//      所以，如果需要改掉模型，方法就是 1. 通过spring.ai.datascope.chat.options.model设置 来改掉默认的模型//                                  2. 或者这样改 dashScopeChatClient.prompt(new Prompt(query, ChatOptions.builder().model("deepseek-r1").build())).call().content()// 3、注意几个类: DashScopeAutoConfiguration、//              DashScopeConnectionProperties（配置前缀是 spring.ai.datascope）、//              DashScopeChatProperties（配置前缀是 spring.ai.datascope.chat）// return dashScopeChatClient.prompt(query).call().content();return dashScopeChatClient.prompt(new Prompt(query, ChatOptions.builder().model("deepseek-r1").build())).call().content();}// http://localhost:8080/helloworld/stream/chat?query=你是谁/*** ChatClient 流式调用*/@GetMapping("/stream/chat")public Flux<String> streamChat(@RequestParam(value = "query", defaultValue = "你好，很高兴认识你，能简单介绍一下自己吗？") String query, HttpServletResponse response) {response.setCharacterEncoding("UTF-8");return dashScopeChatClient.prompt(query).stream().content();}// http://127.0.0.1:8080/helloworld/advisor/chat/123?query=你好，我叫牧生，之后的会话中都带上我的名字// http://127.0.0.1:8080/helloworld/advisor/chat/123?query=我叫什么名字？/*** ChatClient 使用自定义的 Advisor 实现功能增强.* eg:* http://127.0.0.1:8080/helloworld/advisor/chat/123?query=你好，我叫牧生，之后的会话中都带上我的名字* 你好，牧生！很高兴认识你。在接下来的对话中，我会记得带上你的名字。有什么想聊的吗？* http://127.0.0.1:8080/helloworld/advisor/chat/123?query=我叫什么名字？* 你叫牧生呀。有什么事情想要分享或者讨论吗，牧生？*/@GetMapping("/advisor/chat/{id}")public Flux<String> advisorChat(HttpServletResponse response,@PathVariable(value = "id") String id,@RequestParam(value = "query") String query) {response.setCharacterEncoding("UTF-8");return this.dashScopeChatClient.prompt(query)// 因为前面使用Client.Builder构建Client时, 添加了MessageChatMemoryAdvisor切面// 所以, 这里可以添加切面参数.advisors(a -> a.param(CHAT_MEMORY_CONVERSATION_ID_KEY, id).param(CHAT_MEMORY_RETRIEVE_SIZE_KEY, 100)).stream().content();}}

DashScopeChatModelController

import com.alibaba.cloud.ai.dashscope.chat.DashScopeChatOptions;
import jakarta.servlet.http.HttpServletResponse;
import reactor.core.publisher.Flux;import org.springframework.ai.chat.model.ChatModel;
import org.springframework.ai.chat.model.ChatResponse;
import org.springframework.ai.chat.prompt.Prompt;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RestController;@RestController
@RequestMapping("/dashscope/chat-model")
public class DashScopeChatModelController {private static final String DEFAULT_PROMPT = "你好，介绍下你自己吧。";private final ChatModel dashScopeChatModel;public DashScopeChatModelController(ChatModel chatModel) {this.dashScopeChatModel = chatModel;}// http://localhost:8080/dashscope/chat-model/simple/chat/*** 最简单的使用方式，没有任何 LLMs 参数注入。* @return String types.*/@GetMapping("/simple/chat")public String simpleChat() {return dashScopeChatModel.call(new Prompt(DEFAULT_PROMPT, DashScopeChatOptions.builder()// .withModel("qwen2.5-vl-72b-instruct") // 这个会报错, 这应该是模型的问题.withModel("qwen-plus") // 这个正常回答: 你好！我是通义千问，阿里巴巴集团旗下...// .withModel("deepseek-r1")  // 这个正常回答: 您好！我是由中国的深度求索（DeepSeek）公司开发的智能助手DeepSeek-R1。如您有任...// .withModel("deepseek-r1-distill-qwen-1.5b")  // 这个正常回答: 我是DeepSeek-R1，一个由深度求索公司开发的智能助手，我....build())).getResult().getOutput().getContent();}// http://localhost:8080/dashscope/chat-model/stream/chat/*** Stream 流式调用。可以使大模型的输出信息实现打字机效果。* @return Flux<String> types.*/@GetMapping("/stream/chat")public Flux<String> streamChat(HttpServletResponse response) {// 避免返回乱码response.setCharacterEncoding("UTF-8");Flux<ChatResponse> stream = dashScopeChatModel.stream(new Prompt(DEFAULT_PROMPT, DashScopeChatOptions.builder()// .withModel("qwen-plus") // 这个正常回答: 你好！我是通义千问，阿里巴巴集团旗下...// .withModel("deepseek-r1")  // 这个正常回答: 您好！我是由中国的深度求索（DeepSeek）公司开发的智能助手DeepSeek-R1。如您有任....withModel("deepseek-r1-distill-qwen-1.5b")  // /这个正常回答: 我是DeepSeek-R1，....build()));return stream.map(resp -> resp.getResult().getOutput().getContent());}// http://localhost:8080/dashscope/chat-model/custom/chat/*** 使用编程方式自定义 LLMs ChatOptions 参数， {@link com.alibaba.cloud.ai.dashscope.chat.DashScopeChatOptions}* 优先级高于在 application.yml 中配置的 LLMs 参数！*/@GetMapping("/custom/chat")public String customChat() {// 不指定模型, 将会使用默认的  spring.ai.dashscope.chat.options.model 配置的模型（默认是qwen-plus）DashScopeChatOptions customOptions = DashScopeChatOptions.builder().withTopP(0.7).withTopK(50).withTemperature(0.8).build();return dashScopeChatModel.call(new Prompt(DEFAULT_PROMPT, customOptions)).getResult().getOutput().getContent();}}

图解spring AI的结构

deepseek

这节示例是使用OpenAi对接deepseek，跟dashscope没关系（所以这个依赖可以不用引入），需要引入spring-ai-openai-spring-boot-starter。如果同时引入spring-ai-alibaba-starter就需要注意 注入的ChatModel的具体实现。

pom.xml

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0"xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd"><modelVersion>4.0.0</modelVersion><parent><artifactId>spring-boot-dependencies</artifactId><groupId>org.springframework.boot</groupId><version>3.4.2</version></parent><groupId>org.example</groupId><artifactId>demo-alibailian</artifactId><version>1.0-SNAPSHOT</version><properties><java.version>17</java.version><maven.compiler.source>17</maven.compiler.source><maven.compiler.target>17</maven.compiler.target></properties><dependencies><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-web</artifactId></dependency><dependency><groupId>com.alibaba</groupId><artifactId>dashscope-sdk-java</artifactId><version>2.18.2</version><exclusions>&lt;!&ndash; 排除slf4j-simple依赖 &ndash;&gt;<exclusion><groupId>org.slf4j</groupId><artifactId>slf4j-simple</artifactId></exclusion></exclusions></dependency><dependency><groupId>com.alibaba.cloud.ai</groupId><artifactId>spring-ai-alibaba-starter</artifactId><version>1.0.0-M5.1</version></dependency><!-- 引入spring-ai openai的启动依赖 --><dependency><groupId>org.springframework.ai</groupId><artifactId>spring-ai-openai-spring-boot-starter</artifactId><version>1.0.0-M5</version></dependency></dependencies><repositories><repository><id>spring-milestones</id><name>Spring Milestones</name><url>https://repo.spring.io/milestone</url><snapshots><enabled>false</enabled></snapshots></repository></repositories></project>

application.yml

spring:ai:# 百炼的配置dashscope:api-key: sk-~~~# 百炼的ApiKeychat:options:model: qwen-plus# 可兼容openai的配置openai:api-key: sk-~~~# deepseek的ApiKeybase-url: https://api.deepseek.comchat:options:model: deepseek-chatembedding:enabled: false

DeepSeekChatClientController

@RestController
public class DeepSeekChatClientController {private static final String DEFAULT_PROMPT = "你好，介绍下你自己！";private final ChatModel chatModel;private final ChatClient deepSeekChatClient;// 【注入的是 OpenAiChatModel, 而这里配置的是与openai兼容的deepseek。//  OpenAiAutoConfiguration 中定义了 OpenAiChatModel。//  ChatModel（这个由各个spring-ai模块包自己实现，并通过自动配置类提供出来）//  -> ChatClient.Builder（这个在 ClientAutoConfiguration 中定义为多例bean，它需要1个 ChatModel）//  -> ChatClient（使用 ChatClient.Builder 构建，其实现为 DefaultChatClient）】public DeepSeekChatClientController (OpenAiChatModel chatModel) {this.chatModel = chatModel;// 这里实际上就是在模拟 ClientAutoConfiguration 中定义 ChatClient.Builder多例bean 类似this.deepSeekChatClient = ChatClient.builder(chatModel).defaultAdvisors(new MessageChatMemoryAdvisor(new InMemoryChatMemory()))// 实现 Logger 的 Advisor.defaultAdvisors(new SimpleLoggerAdvisor())// 设置 ChatClient 中 ChatModel 的 Options 参数.defaultOptions(OpenAiChatOptions.builder().temperature(0.7d).build()).build();}// http://localhost:8080/ai/customOptions/*** 使用自定义参数调用DeepSeek模型** @return ChatResponse 包含模型响应结果的封装对象* @apiNote 当前硬编码指定模型为deepseek-chat，温度参数0.7以平衡生成结果的创造性和稳定性*/@GetMapping(value = "/ai/customOptions")public ChatResponse testDeepSeekCustomOptions () {return this.deepSeekChatClient.prompt(DEFAULT_PROMPT).call().chatResponse();}// http://localhost:8080/ai/generate/*** 执行默认提示语的AI生成请求** @return Map 包含生成结果的键值对，格式为{ "generation": 响应内容 }*/@GetMapping("/ai/generate")public Map<String, Object> testEasyChat () {return Map.of("generation", this.deepSeekChatClient.prompt(DEFAULT_PROMPT).call().content());}// http://localhost:8080/ai/stream/*** 流式生成接口 - 支持实时获取生成过程的分块响应** @return Flux<ChatResponse> 响应式流对象，包含分块的模型响应数据* @see Flux 基于Project Reactor的响应式流对象*/@GetMapping("/ai/stream")public Flux<String> testDeepSeekGenerateWithStream (HttpServletResponse response) {response.setCharacterEncoding("UTF-8");// 1、DefaultChatClientRequestSpec 构造方法中添加的 StreamAroundAdvisor 触发了请求的执行//    调用 chatModel.stream(Prompt)，即 chatModel#stream(Prompt)，这里的chatModel就是OpenAiChatModel的实现// 2、所以具体怎么解析响应的 需要到 OpenAiChatModel#stream(Prompt) 方法中查看// 3、所以，其实就发现 chatClient还绕了一圈，然后再去调用chatModel，绕一圈的目的是为了增强功能return deepSeekChatClient.prompt(DEFAULT_PROMPT).stream().chatResponse().map(chatResponse -> chatResponse.getResult().getOutput().getContent());}}

DeepSeekChatModelController

@RestController
public class DeepSeekChatModelController {private static final String DEFAULT_PROMPT = "你好，介绍下你自己吧。";private final ChatModel deepSeekChatModel;// 注入的是 OpenAiChatModelpublic DeepSeekChatModelController (OpenAiChatModel chatModel) {this.deepSeekChatModel = chatModel;}// http://localhost:8080/simple/chat/*** 最简单的使用方式，没有任何 LLMs 参数注入。** @return String types.*/@GetMapping("/simple/chat")public String simpleChat () {return deepSeekChatModel.call(new Prompt(DEFAULT_PROMPT)).getResult().getOutput().getContent();}// http://localhost:8080/stream/chat/*** Stream 流式调用。可以使大模型的输出信息实现打字机效果。** @return Flux<String> types.*/@GetMapping("/stream/chat")public Flux<String> streamChat (HttpServletResponse response) {response.setCharacterEncoding("UTF-8");Flux<ChatResponse> stream = deepSeekChatModel.stream(new Prompt(DEFAULT_PROMPT));return stream.map(resp -> resp.getResult().getOutput().getContent());}// http://localhost:8080/custom/chat/*** 使用编程方式自定义 LLMs ChatOptions 参数， {OpenAIChatOption}* 优先级高于在 application.yml 中配置的 LLMs 参数！*/@GetMapping("/custom/chat")public String customChat () {// 指定的模型 将覆盖 chatModel 中配置的模型// 模型名优先级顺序 Prompt中的options > client配置的options > chatModel中的options(通过配置文件配置的)OpenAiChatOptions customOptions = OpenAiChatOptions.builder().model("deepseek-reasoner").temperature(0.8d).build();return deepSeekChatModel.call(new Prompt(DEFAULT_PROMPT, customOptions)).getResult().getOutput().getContent();}
}

dashscope文生图

pom.xml

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0"xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd"><modelVersion>4.0.0</modelVersion><parent><artifactId>spring-boot-dependencies</artifactId><groupId>org.springframework.boot</groupId><version>3.4.2</version></parent><groupId>org.example</groupId><artifactId>demo-alibailian</artifactId><version>1.0-SNAPSHOT</version><properties><java.version>17</java.version><maven.compiler.source>17</maven.compiler.source><maven.compiler.target>17</maven.compiler.target></properties><dependencies><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-web</artifactId></dependency><dependency><groupId>com.alibaba</groupId><artifactId>dashscope-sdk-java</artifactId><version>2.18.2</version><exclusions><!-- 排除slf4j-simple依赖 --><exclusion><groupId>org.slf4j</groupId><artifactId>slf4j-simple</artifactId></exclusion></exclusions></dependency><dependency><groupId>com.alibaba.cloud.ai</groupId><artifactId>spring-ai-alibaba-starter</artifactId><version>1.0.0-M5.1</version></dependency></dependencies><repositories><repository><id>spring-milestones</id><name>Spring Milestones</name><url>https://repo.spring.io/milestone</url><snapshots><enabled>false</enabled></snapshots></repository></repositories></project>

application.yml

spring:ai:# 百炼的配置dashscope:api-key: sk-~~~# 百炼的ApiKey

DashScopeImageController

可以看出来，spring ai alibaba其实就是实现了spring ai core中的接口

import java.io.IOException;
import java.io.InputStream;
import java.net.URI;
import java.net.URL;import jakarta.servlet.http.HttpServletResponse;import org.springframework.ai.image.ImageModel;
import org.springframework.ai.image.ImagePrompt;
import org.springframework.ai.image.ImageResponse;
import org.springframework.http.MediaType;
import org.springframework.web.bind.annotation.GetMapping;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RequestParam;
import org.springframework.web.bind.annotation.RestController;@RestController
@RequestMapping("/example")
public class DashScopeImageController {private final ImageModel imageModel;private static final String DEFAULT_PROMPT = "为人工智能生成一张富有科技感的图片！";public DashScopeImageController(ImageModel imageModel) {this.imageModel = imageModel;}// http://localhost:8080/example/image?text=生成1张小怪兽打奥特曼的图片@GetMapping("/image")public void image(@RequestParam(value = "text", defaultValue = DEFAULT_PROMPT) String text, HttpServletResponse response) {ImageResponse imageResponse = imageModel.call(new ImagePrompt(text));String imageUrl = imageResponse.getResult().getOutput().getUrl();System.out.println(imageUrl);try {URL url = URI.create(imageUrl).toURL();InputStream in = url.openStream();response.setHeader("Content-Type", MediaType.IMAGE_PNG_VALUE);response.getOutputStream().write(in.readAllBytes());response.getOutputStream().flush();} catch (IOException e) {response.setStatus(HttpServletResponse.SC_INTERNAL_SERVER_ERROR);}}}

dashscope语音

pom.xml

<?xml version="1.0" encoding="UTF-8"?>
<project xmlns="http://maven.apache.org/POM/4.0.0"xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd"><modelVersion>4.0.0</modelVersion><parent><artifactId>spring-boot-dependencies</artifactId><groupId>org.springframework.boot</groupId><version>3.4.2</version></parent><groupId>org.example</groupId><artifactId>demo-alibailian</artifactId><version>1.0-SNAPSHOT</version><properties><java.version>17</java.version><maven.compiler.source>17</maven.compiler.source><maven.compiler.target>17</maven.compiler.target></properties><dependencies><dependency><groupId>org.springframework.boot</groupId><artifactId>spring-boot-starter-web</artifactId></dependency><dependency><groupId>com.alibaba</groupId><artifactId>dashscope-sdk-java</artifactId><version>2.18.2</version><exclusions><!-- 排除slf4j-simple依赖 --><exclusion><groupId>org.slf4j</groupId><artifactId>slf4j-simple</artifactId></exclusion></exclusions></dependency><dependency><groupId>com.alibaba.cloud.ai</groupId><artifactId>spring-ai-alibaba-starter</artifactId><version>1.0.0-M5.1</version></dependency></dependencies><repositories><repository><id>spring-milestones</id><name>Spring Milestones</name><url>https://repo.spring.io/milestone</url><snapshots><enabled>false</enabled></snapshots></repository></repositories></project>

application.yml

spring:ai:# 百炼的配置dashscope:api-key: sk-~~~# 百炼的ApiKey

TTSController

@RestController
@RequestMapping("/ai/tts")
public class TTSController /*implements ApplicationRunner*/ {private final SpeechSynthesisModel speechSynthesisModel;private static final String TEXT = "白日依山尽，黄河入海流。";private static final String FILE_PATH = System.getProperty("user.dir") + "/file/";public TTSController(SpeechSynthesisModel speechSynthesisModel) {this.speechSynthesisModel = speechSynthesisModel;}// http://localhost:8080/ai/tts?text=白日依山尽，黄河入海流。@GetMappingpublic void tts(@RequestParam(value = "text", defaultValue = TEXT) String text) throws IOException {SpeechSynthesisResponse response = speechSynthesisModel.call(new SpeechSynthesisPrompt(text));File file = new File(FILE_PATH + "/output.mp3");try (FileOutputStream fos = new FileOutputStream(file)) {ByteBuffer byteBuffer = response.getResult().getOutput().getAudio();fos.write(byteBuffer.array());}catch (IOException e) {throw new IOException(e.getMessage());}}@GetMapping("/stream")public void streamTTS(@RequestParam(value = "text", defaultValue = TEXT) String text) {Flux<SpeechSynthesisResponse> response = speechSynthesisModel.stream(new SpeechSynthesisPrompt(text));CountDownLatch latch = new CountDownLatch(1);File file = new File(FILE_PATH + "/output-stream.mp3");try (FileOutputStream fos = new FileOutputStream(file)) {response.doFinally(signal -> latch.countDown()).subscribe(synthesisResponse -> {ByteBuffer byteBuffer = synthesisResponse.getResult().getOutput().getAudio();byte[] bytes = new byte[byteBuffer.remaining()];byteBuffer.get(bytes);try {fos.write(bytes);}catch (IOException e) {throw new RuntimeException(e);}});latch.await();}catch (IOException | InterruptedException e) {throw new RuntimeException(e);}}/*@Overridepublic void run(ApplicationArguments args) {File file = new File(FILE_PATH);if (!file.exists()) {file.mkdirs();}}@PreDestroypublic void destroy() throws IOException {FileUtils.deleteDirectory(new File(FILE_PATH));}
*/
}

STTController

@RestController
@RequestMapping("/ai/stt")
public class STTController {private final AudioTranscriptionModel transcriptionModel;private static final Logger log = LoggerFactory.getLogger(STTController.class);private static final String DEFAULT_MODEL_1 = "sensevoice-v1";private static final String DEFAULT_MODEL_2 = "paraformer-realtime-v2";private static final String DEFAULT_MODEL_3 = "paraformer-v2";private static final String AUDIO_RESOURCES_URL = "https://dashscope.oss-cn-beijing.aliyuncs.com/samples/audio/paraformer/hello_world_female2.wav";private final ScheduledExecutorService scheduler = Executors.newScheduledThreadPool(1);public STTController(AudioTranscriptionModel transcriptionModel) {this.transcriptionModel = transcriptionModel;}@GetMappingpublic String stt() throws MalformedURLException {AudioTranscriptionResponse response = transcriptionModel.call(new AudioTranscriptionPrompt(new UrlResource(AUDIO_RESOURCES_URL),DashScopeAudioTranscriptionOptions.builder().withModel(DEFAULT_MODEL_1).build()));return response.getResult().getOutput();}@GetMapping("/stream")public String streamSTT() {CountDownLatch latch = new CountDownLatch(1);StringBuilder stringBuilder = new StringBuilder();Flux<AudioTranscriptionResponse> response = transcriptionModel.stream(new AudioTranscriptionPrompt(new FileSystemResource("spring-ai-alibaba-audio-example/dashscope-audio/src/main/resources/stt/count.pcm"),DashScopeAudioTranscriptionOptions.builder().withModel(DEFAULT_MODEL_2).withSampleRate(16000).withFormat(DashScopeAudioTranscriptionOptions.AudioFormat.PCM).withDisfluencyRemovalEnabled(false).build()));response.doFinally(signal -> latch.countDown()).subscribe(resp -> stringBuilder.append(resp.getResult().getOutput()));try {latch.await();}catch (InterruptedException e) {throw new RuntimeException(e);}return stringBuilder.toString();}@GetMapping("/async")public String asyncSTT() {StringBuilder stringBuilder = new StringBuilder();CountDownLatch latch = new CountDownLatch(1);try {AudioTranscriptionResponse submitResponse = transcriptionModel.asyncCall(new AudioTranscriptionPrompt(new UrlResource(AUDIO_RESOURCES_URL),DashScopeAudioTranscriptionOptions.builder().withModel(DEFAULT_MODEL_3).build()));DashScopeAudioTranscriptionApi.Response.Output submitOutput = Objects.requireNonNull(submitResponse.getMetadata().get("output"));String taskId = submitOutput.taskId();scheduler.scheduleAtFixedRate(() -> checkTaskStatus(taskId, stringBuilder, latch), 0, 1, TimeUnit.SECONDS);latch.await();}catch (MalformedURLException e) {throw new DashScopeException("Error in URL format: " + e.getMessage());}catch (InterruptedException e) {Thread.currentThread().interrupt();throw new DashScopeException("Thread was interrupted: " + e.getMessage());}finally {scheduler.shutdown();}return stringBuilder.toString();}private void checkTaskStatus(String taskId, StringBuilder stringBuilder, CountDownLatch latch) {try {AudioTranscriptionResponse fetchResponse = transcriptionModel.fetch(taskId);DashScopeAudioTranscriptionApi.Response.Output fetchOutput =Objects.requireNonNull(fetchResponse.getMetadata().get("output"));DashScopeAudioTranscriptionApi.TaskStatus taskStatus = fetchOutput.taskStatus();if (taskStatus.equals(DashScopeAudioTranscriptionApi.TaskStatus.SUCCEEDED)) {stringBuilder.append(fetchResponse.getResult().getOutput());latch.countDown();}else if (taskStatus.equals(DashScopeAudioTranscriptionApi.TaskStatus.FAILED)) {log.warn("Transcription failed.");latch.countDown();}}catch (Exception e) {latch.countDown();throw new RuntimeException("Error occurred while checking task status: " + e.getMessage());}}}