Spring AI + Ollama简单使用

一、环境准备

1. 前置条件

1. 本地安装 Ollama，启动服务（默认地址：http://localhost:11434）
2. 拉取任意模型（示例用 qwen3.5:4b，轻量快速）

   docker exec -it ollama ollama run qwen3.5:4b

3. SpringBoot 4.x 项目

2. Maven 依赖 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 https://maven.apache.org/xsd/maven-4.0.0.xsd"> <modelVersion>4.0.0</modelVersion> <parent> <groupId>org.springframework.boot</groupId> <artifactId>spring-boot-starter-parent</artifactId> <version>4.1.0</version> <relativePath/> <!-- lookup parent from repository --> </parent> <groupId>com.ybw</groupId> <artifactId>spring-ai-demo</artifactId> <version>1.0.0</version> <name>spring-ai-demo</name> <description>spring-ai-demo</description> <modules> <module>quick-start</module> </modules> <packaging>pom</packaging> <properties> <java.version>21</java.version> <spring-ai.version>2.0.0</spring-ai.version> </properties> <dependencies> <dependency> <groupId>org.springframework.boot</groupId> <artifactId>spring-boot-starter-webmvc</artifactId> </dependency> <dependency> <groupId>org.springframework.ai</groupId> <artifactId>spring-ai-starter-model-ollama</artifactId> </dependency> <dependency> <groupId>org.projectlombok</groupId> <artifactId>lombok</artifactId> <optional>true</optional> </dependency> <dependency> <groupId>org.springframework.boot</groupId> <artifactId>spring-boot-starter-webmvc-test</artifactId> <scope>test</scope> </dependency> </dependencies> <dependencyManagement> <dependencies> <dependency> <groupId>org.springframework.ai</groupId> <artifactId>spring-ai-bom</artifactId> <version>${spring-ai.version}</version> <type>pom</type> <scope>import</scope> </dependency> </dependencies> </dependencyManagement> <build> <plugins> <plugin> <groupId>org.springframework.boot</groupId> <artifactId>spring-boot-maven-plugin</artifactId> <configuration> <excludes> <exclude> <groupId>org.projectlombok</groupId> <artifactId>lombok</artifactId> </exclude> </excludes> </configuration> </plugin> <plugin> <groupId>org.apache.maven.plugins</groupId> <artifactId>maven-compiler-plugin</artifactId> <executions> <execution> <id>default-compile</id> <phase>compile</phase> <goals> <goal>compile</goal> </goals> <configuration> <annotationProcessorPaths> <path> <groupId>org.projectlombok</groupId> <artifactId>lombok</artifactId> </path> </annotationProcessorPaths> </configuration> </execution> <execution> <id>default-testCompile</id> <phase>test-compile</phase> <goals> <goal>testCompile</goal> </goals> <configuration> <annotationProcessorPaths> <path> <groupId>org.projectlombok</groupId> <artifactId>lombok</artifactId> </path> </annotationProcessorPaths> </configuration> </execution> </executions> </plugin> </plugins> </build> </project>

二、配置文件 application.yml

spring: application: name: spring-ai-demo ai: ollama: base-url: http://127.0.0.1:11434 chat: model: qwen3.5:4b # 模型 temperature: 0.7 # 温度，数值越高，输出结果越随机，数值越低，输出结果越一致。范围0-1 top-p: 0.7 # 概率 think: true # 是否思考

温度

temperature 取值范围：0 ~ 1，核心逻辑：数值越大，模型随机性、创造性越强；数值越小，输出越确定、严谨、重复度低。

三、配置类 + 测试 Demo

配置 ChatClient

package com.ybw.config; import org.springframework.ai.chat.client.ChatClient; import org.springframework.context.annotation.Bean; import org.springframework.context.annotation.Configuration; /** * 配置 ChatClient * * @author ybw * @version V1.0 * @className ChatClientConfig * @date 2026/6/29 **/ @Configuration public class ChatClientConfig { /** * 注入 ChatClient，SpringAI 自动装配 OllamaChatModel * * @param builder 构建 ChatClient * @methodName: chatClient * @return: org.springframework.ai.chat.client.ChatClient * @author: ybw * @date: 2026/6/29 **/ @Bean ChatClient chatClient(ChatClient.Builder builder) { return builder.build(); } }

单元测试

无思考过程

package com.ybw; import jakarta.annotation.Resource; import lombok.extern.slf4j.Slf4j; import org.junit.jupiter.api.Test; import org.springframework.ai.chat.client.ChatClient; import org.springframework.boot.test.context.SpringBootTest; /** * 测试ollama * * @author ybw * @version V1.0 * @className TestOllama * @date 2026/6/29 **/ @SpringBootTest @Slf4j public class TestOllama { @Resource private ChatClient chatClient; /** * 测试ollama:简单聊天demo,打印出聊天结果 */ @Test public void testOllama() { // 1. 单次简单提问 String prompt = "请写一首描述清晨的诗"; log.info("===== 用户提问：{} =====", prompt); // 同步调用 ollama，获取完整返回文本 String response = chatClient.prompt() .user(prompt) .call() .content(); // 打印大模型回答 log.info("===== Ollama 返回结果 ====="); log.info(response); log.info("==========================="); // 2. 多轮对话示例（携带上下文） log.info("===== 多轮对话测试 ====="); String multiRoundRes = chatClient.prompt() .user("什么是Java？") .system("你是简洁的编程讲师，回答不超过两句话") .call() .content(); log.info(multiRoundRes); } /** * 测试ollama:流式输出聊天结果 */ @Test public void testOllamaStream() { // 流式响应，逐段打印 chatClient.prompt() .user("写一段100字的春日短文") .stream() .content() .doOnNext(System.out::print) .blockLast(); } }

• 流式输出（逐行实时打印，类似打字机效果）

有思考过程

package com.ybw; import jakarta.annotation.Resource; import lombok.extern.slf4j.Slf4j; import org.junit.jupiter.api.Test; import org.springframework.ai.chat.client.ChatClient; import org.springframework.ai.chat.messages.AssistantMessage; import org.springframework.ai.chat.model.ChatResponse; import org.springframework.boot.test.context.SpringBootTest; import reactor.core.publisher.Flux; import java.util.Map; /** * 测试ollama思考过程 * * @author ybw * @version V1.0 * @className TestThinkOllama * @date 2026/6/29 **/ @SpringBootTest @Slf4j public class TestThinkOllama { @Resource private ChatClient chatClient; /** * 测试ollama:思考过程 */ @Test public void testThinkOllama() { //1. 创建一个 Prompt String prompt = "java是什么？"; log.info("===== 用户提问：{} =====", prompt); //2. 调用 ollama ChatResponse chatResponse = chatClient.prompt() .user(prompt) .call() .chatResponse(); // 3. 关键修改：获取完整输出内容 AssistantMessage assistantMessage = chatResponse.getResult().getOutput(); // 3.1 获取思考过程（从 metadata 中提取） Map<String, Object> metadata = assistantMessage.getMetadata(); String thinking = metadata.get("thinking").toString(); log.info("===== 思考过程 ====="); log.info(thinking); // 3.2 获取最终答案 String finalAnswer = assistantMessage.getText(); log.info("===== 最终答案 ====="); log.info(finalAnswer); } /** * 测试ollama:流式输出聊天结果 */ @Test public void testThinkOllamaStream() { //1. 最终结果存储 StringBuilder finishThinkingStr = new StringBuilder(); StringBuilder finishAnswerStr = new StringBuilder(); //1. 流式响应，逐段打印 Flux<ChatResponse> chatResponseFlux = chatClient.prompt() .user("你好") .stream() .chatResponse(); // 2. 逐段打印 chatResponseFlux.doOnNext(chatResponse -> { // 2.1 获取完整输出内容 AssistantMessage assistantMessage = chatResponse.getResult().getOutput(); // 2.2 获取思考过程（从 metadata 中提取） Map<String, Object> metadata = assistantMessage.getMetadata(); Object thinking = metadata.get("thinking"); if (thinking != null) { String thinkingStr = thinking.toString(); // 灰色显示思考过程 System.out.print("\u001B[90m" + thinkingStr + "\u001B[0m"); finishThinkingStr.append(thinkingStr); } else { //只有刚开始打印的时候，换一次行 if (finishAnswerStr.isEmpty()) { System.out.println(); } // 2.3 获取最终答案 String finalAnswer = assistantMessage.getText(); System.out.print(finalAnswer); finishAnswerStr.append(finalAnswer); } }).doOnComplete(() -> { System.out.println(); log.info("思考:{}", finishThinkingStr); log.info("回复:{}", finishAnswerStr); }) // 等待流结束 .blockLast(); } }