如何在Java中仅通过Bearer API token调用Gemini Pro模型

如何在Java中仅通过（Bearer）API token作为唯一认证方式，向Gemini Pro模型发送“你能说‘你好’吗？”的请求？

• 可通过命令 gcloud auth application-default print-access-token 获取Bearer API token
• 已实现常规认证方式调用Gemini Pro模型的Java代码，但需改为仅用Bearer token认证的方式
• 假设获取到的Bearer token可在未登录谷歌账号的设备上直接使用，无需账号密码

方式一：直接调用Vertex AI REST API（无需SDK） #

不依赖Vertex AI Java SDK，直接通过HTTP请求调用API，手动将Bearer token加入请求头：

import java.io.IOException;
import java.net.URI;
import java.net.http.HttpClient;
import java.net.http.HttpRequest;
import java.net.http.HttpResponse;

public class GeminiTokenAuth {
    public static void main(String[] args) throws IOException, InterruptedException {
        // 替换为你的实际参数
        String bearerToken = "你的Bearer API token";
        String projectId = "你的谷歌云项目ID";
        String location = "us-central1";
        String modelName = "gemini-pro";

        // 构建API请求地址
        String apiUrl = String.format(
            "https://%s-aiplatform.googleapis.com/v1/projects/%s/locations/%s/publishers/google/models/%s:generateContent",
            location, projectId, location, modelName
        );

        // 构建请求体
        String requestBody = """
            {
                "contents": [
                    {
                        "parts": [{"text": "你能说‘你好’吗？"}]
                    }
                ]
            }
            """;

        // 发送HTTP请求
        HttpClient client = HttpClient.newHttpClient();
        HttpRequest request = HttpRequest.newBuilder()
            .uri(URI.create(apiUrl))
            .header("Authorization", "Bearer " + bearerToken)
            .header("Content-Type", "application/json")
            .POST(HttpRequest.BodyPublishers.ofString(requestBody))
            .build();

        HttpResponse<String> response = client.send(request, HttpResponse.BodyHandlers.ofString());

        // 输出结果
        System.out.println("响应状态码: " + response.statusCode());
        System.out.println("响应内容: " + response.body());
    }
}

方式二：修改Vertex AI SDK的认证配置 #

若想继续使用Vertex AI Java SDK，可通过自定义Credentials注入Bearer token：

import com.google.auth.Credentials;
import com.google.cloud.vertexai.VertexAI;
import com.google.cloud.vertexai.api.GenerateContentResponse;
import com.google.cloud.vertexai.generativeai.preview.GenerativeModel;
import com.google.cloud.vertexai.generativeai.preview.ResponseHandler;
import java.io.IOException;
import java.net.URI;
import java.util.Collections;
import java.util.List;
import java.util.Map;

public class GeminiSdkTokenAuth {
    public static void main(String[] args) throws IOException {
        // 替换为你的实际参数
        String bearerToken = "你的Bearer API token";
        String projectId = "你的谷歌云项目ID";
        String location = "us-central1";
        String modelName = "gemini-pro";

        // 自定义Credentials，注入Bearer token
        Credentials customCredentials = new Credentials() {
            @Override
            public String getAuthenticationType() {
                return "Bearer";
            }

            @Override
            public Map<String, List<String>> getRequestMetadata(URI uri) throws IOException {
                return Collections.singletonMap(
                    "Authorization",
                    Collections.singletonList("Bearer " + bearerToken)
                );
            }

            @Override
            public boolean hasRequestMetadata() {
                return true;
            }

            @Override
            public boolean hasRequestMetadataOnly() {
                return true;
            }

            @Override
            public void refresh() throws IOException {}
        };

        // 使用自定义Credentials初始化VertexAI客户端并调用模型
        try (VertexAI vertexAI = new VertexAI(projectId, location, customCredentials)) {
            GenerativeModel model = new GenerativeModel(modelName, vertexAI);
            GenerateContentResponse response = model.generateContent("你能说‘你好’吗？");
            System.out.println(ResponseHandler.getText(response));
        }
    }
}

• Bearer token存在有效期，过期后需重新获取
• 确保你的token拥有调用Vertex AI Gemini Pro模型的权限
• 使用REST API时，需保证请求体格式符合Vertex AI API的规范

内容的提问来源于stack exchange，提问作者a.t.