Spring AI + Ollama + Qwen3.5，快速搭建一个ai应用

零成本不联网一台普通办公本就能跑。你需要准备JDK 178GB 以上内存4G以上显卡虽然ollama可以用cpu但是token吞吐太慢了第一步装 Ollama拉模型ollama.com 下载安装然后ollama pull qwen3.5:0.8b500MB几分钟下完。验证一下ollama run qwen3.5:0.8b 你好能回你继续。如果出现错误参考我上一篇文章第二步版本别搞错我一开始就搞错版本了总是包找不到Spring BootSpring AI3.4.x1.0.x3.5.x1.1.x4.0.x2.0.x本文用Spring Boot 3.4.4 Spring AI 1.0.8。第三步贴代码pom.xmlparent groupIdorg.springframework.boot/groupId artifactIdspring-boot-starter-parent/artifactId version3.4.4/version /parent properties java.version17/java.version spring-ai.version1.0.8/spring-ai.version /properties dependencyManagement dependencies dependency groupIdorg.springframework.ai/groupId artifactIdspring-ai-bom/artifactId version${spring-ai.version}/version typepom/type scopeimport/scope /dependency /dependencies /dependencyManagement dependencies dependency groupIdorg.springframework.boot/groupId artifactIdspring-boot-starter-web/artifactId /dependency dependency groupIdorg.springframework.ai/groupId artifactIdspring-ai-ollama-spring-boot-starter/artifactId /dependency /dependenciesapplication.ymlspring: ai: ollama: base-url: http://localhost:11434 #默认端口 chat: model: qwen3.5:0.8b options: temperature: 0.7AiConfig.javaConfiguration public class AiConfig { Bean public ChatClient chatClient(ChatClient.Builder builder) { return builder .defaultSystem(用简洁中文回答。) //指令 .build(); } }ChatController.javaRestController public class ChatController { private final ChatClient chatClient; public ChatController(ChatClient chatClient) { this.chatClient chatClient; } GetMapping(/chat) public String chat(RequestParam String msg) { return chatClient.prompt().user(msg).call().content(); } }业务代码就一行。StreamController.java流式RestController public class StreamController { private final ChatClient chatClient; public StreamController(ChatClient chatClient) { this.chatClient chatClient; } GetMapping(value /chat/stream, produces MediaType.TEXT_EVENT_STREAM_VALUE) public FluxString stream(RequestParam String msg) { return chatClient.prompt().user(msg).stream().content(); } }第四步跑起来mvn spring-boot:run日志看到这行就对了Ollama chat model initialized with model: qwen3.5:0.8b第五步测试# 对话 curl http://localhost:8080/chat?msg用Java写个冒泡排序 # 翻译 curl http://localhost:8080/chat?msg把今天天气真好翻译成英文 # 流式浏览器打开看逐字输出 http://localhost:8080/chat/stream?msg写一首五言绝句踩坑速查报错解决Connection refused: 11434Ollama 没启动model not foundollama pull qwen3.5:0.8bNoClassDefFoundError版本没对上回第二步查表响应慢正常CPU 推理 0.8B 就这速度快来试一试把快速搭建属于自己的AI应用项目结构ai-demo/ ├── pom.xml ├── src/main/java/com/example/demo/ │ ├── DemoApplication.java │ ├── config/AiConfig.java │ └── controller/ │ ├── ChatController.java │ └── StreamController.java └── src/main/resources/application.yml