関数呼び出しを使用してコンテンツを生成する

このサンプルでは、関数宣言を使用して Gemini マルチモーダルで生成されるコンテンツに影響を与える方法を示します。

詳細

このコードサンプルを含む詳細なドキュメントについては、以下をご覧ください。

• 関数呼び出しのリファレンス
• 関数呼び出しの概要

コードサンプル

Go

このサンプルを試す前に、Vertex AI クイックスタート: クライアント ライブラリの使用にある Go の設定手順を完了してください。 詳細については、Vertex AI Go API のリファレンス ドキュメントをご覧ください。

Vertex AI に対する認証を行うには、アプリケーションのデフォルト認証情報を設定します。詳細については、ローカル開発環境の認証を設定するをご覧ください。

import (
	"context"
	"fmt"
	"io"

	genai "google.golang.org/genai"
)

// generateWithFuncCall shows how to submit a prompt and a function declaration to the model,
// allowing it to suggest a call to the function to fetch external data. Returning this data
// enables the model to generate a text response that incorporates the data.
func generateWithFuncCall(w io.Writer) error {
	ctx := context.Background()

	client, err := genai.NewClient(ctx, &genai.ClientConfig{
		HTTPOptions: genai.HTTPOptions{APIVersion: "v1"},
	})
	if err != nil {
		return fmt.Errorf("failed to create genai client: %w", err)
	}

	weatherFunc := &genai.FunctionDeclaration{
		Description: "Returns the current weather in a location.",
		Name:        "getCurrentWeather",
		Parameters: &genai.Schema{
			Type: "object",
			Properties: map[string]*genai.Schema{
				"location": {Type: "string"},
			},
			Required: []string{"location"},
		},
	}
	config := &genai.GenerateContentConfig{
		Tools: []*genai.Tool{
			{FunctionDeclarations: []*genai.FunctionDeclaration{weatherFunc}},
		},
		Temperature: genai.Ptr(float32(0.0)),
	}

	modelName := "gemini-2.5-flash"
	contents := []*genai.Content{
		{Parts: []*genai.Part{
			{Text: "What is the weather like in Boston?"},
		},
			Role: "user"},
	}

	resp, err := client.Models.GenerateContent(ctx, modelName, contents, config)
	if err != nil {
		return fmt.Errorf("failed to generate content: %w", err)
	}

	var funcCall *genai.FunctionCall
	for _, p := range resp.Candidates[0].Content.Parts {
		if p.FunctionCall != nil {
			funcCall = p.FunctionCall
			fmt.Fprint(w, "The model suggests to call the function ")
			fmt.Fprintf(w, "%q with args: %v\n", funcCall.Name, funcCall.Args)
			// Example response:
			// The model suggests to call the function "getCurrentWeather" with args: map[location:Boston]
		}
	}
	if funcCall == nil {
		return fmt.Errorf("model did not suggest a function call")
	}

	// Use synthetic data to simulate a response from the external API.
	// In a real application, this would come from an actual weather API.
	funcResp := &genai.FunctionResponse{
		Name: "getCurrentWeather",
		Response: map[string]any{
			"location":         "Boston",
			"temperature":      "38",
			"temperature_unit": "F",
			"description":      "Cold and cloudy",
			"humidity":         "65",
			"wind":             `{"speed": "10", "direction": "NW"}`,
		},
	}

	// Return conversation turns and API response to complete the model's response.
	contents = []*genai.Content{
		{Parts: []*genai.Part{
			{Text: "What is the weather like in Boston?"},
		},
			Role: "user"},
		{Parts: []*genai.Part{
			{FunctionCall: funcCall},
		}},
		{Parts: []*genai.Part{
			{FunctionResponse: funcResp},
		}},
	}

	resp, err = client.Models.GenerateContent(ctx, modelName, contents, config)
	if err != nil {
		return fmt.Errorf("failed to generate content: %w", err)
	}

	respText := resp.Text()

	fmt.Fprintln(w, respText)

	// Example response:
	// The weather in Boston is cold and cloudy with a temperature of 38 degrees Fahrenheit. The humidity is ...

	return nil
}

Java

このサンプルを試す前に、Vertex AI クイックスタート: クライアント ライブラリの使用にある Java の設定手順を完了してください。詳細については、Vertex AI Java API のリファレンス ドキュメントをご覧ください。

Vertex AI に対する認証を行うには、アプリケーションのデフォルト認証情報を設定します。詳細については、ローカル開発環境の認証を設定するをご覧ください。

import com.google.genai.Client;
import com.google.genai.types.FunctionDeclaration;
import com.google.genai.types.GenerateContentConfig;
import com.google.genai.types.GenerateContentResponse;
import com.google.genai.types.HttpOptions;
import com.google.genai.types.Schema;
import com.google.genai.types.Tool;
import com.google.genai.types.Type;
import java.util.Map;

public class ToolFunctionDescriptionWithText {

  public static void main(String[] args) {
    // TODO(developer): Replace these variables before running the sample.
    String modelId = "gemini-2.5-flash";
    String contents =
        "At Stellar Sounds, a music label, 2024 was a rollercoaster. \"Echoes of the Night,\""
            + " a debut synth-pop album, \n surprisingly sold 350,000 copies, while veteran"
            + " rock band \"Crimson Tide's\" latest, \"Reckless Hearts,\" \n lagged at"
            + " 120,000. Their up-and-coming indie artist, \"Luna Bloom's\" EP, \"Whispers "
            + "of Dawn,\" \n secured 75,000 sales. The biggest disappointment was the "
            + "highly-anticipated rap album \"Street Symphony\" \n only reaching 100,000"
            + " units. Overall, Stellar Sounds moved over 645,000 units this year, revealing"
            + " unexpected \n trends in music consumption.";

    generateContent(modelId, contents);
  }

  // Generates content with text input and function declaration that
  // the model may use to retrieve external data for the response
  public static String generateContent(String modelId, String contents) {
    // Initialize client that will be used to send requests. This client only needs to be created
    // once, and can be reused for multiple requests.
    try (Client client =
        Client.builder()
            .location("global")
            .vertexAI(true)
            .httpOptions(HttpOptions.builder().apiVersion("v1").build())
            .build()) {

      FunctionDeclaration getAlbumSales =
          FunctionDeclaration.builder()
              .name("get_album_sales")
              .description("Gets the number of albums sold")
              // Function parameters are specified in schema format
              .parameters(
                  Schema.builder()
                      .type(Type.Known.OBJECT)
                      .properties(
                          Map.of(
                              "albums",
                              Schema.builder()
                                  .type(Type.Known.ARRAY)
                                  .description("List of albums")
                                  .items(
                                      Schema.builder()
                                          .description("Album and its sales")
                                          .type(Type.Known.OBJECT)
                                          .properties(
                                              Map.of(
                                                  "album_name",
                                                      Schema.builder()
                                                          .type(Type.Known.STRING)
                                                          .description("Name of the music album")
                                                          .build(),
                                                  "copies_sold",
                                                      Schema.builder()
                                                          .type(Type.Known.INTEGER)
                                                          .description("Number of copies sold")
                                                          .build()))
                                          .build()) // End items schema for albums
                                  .build() // End "albums" property schema
                              ))
                      .build()) // End parameters schema
              .build(); // End function declaration

      Tool salesTool = Tool.builder().functionDeclarations(getAlbumSales).build();

      GenerateContentConfig config =
          GenerateContentConfig.builder().tools(salesTool).temperature(0.0f).build();

      GenerateContentResponse response = client.models.generateContent(modelId, contents, config);

      // response.functionCalls() returns an Immutable<FunctionCall>.
      System.out.println(response.functionCalls().get(0));

      return response.functionCalls().toString();
      // Example response:
      // FunctionCall{id=Optional.empty, args=Optional[{albums=[{copies_sold=350000,
      // album_name=Echoes of the Night},
      // {copies_sold=120000, album_name=Reckless Hearts}, {copies_sold=75000, album_name=Whispers
      // of Dawn},
      // {album_name=Street Symphony, copies_sold=100000}]}], name=Optional[get_album_sales]}
    }
  }
}

Node.js

このサンプルを試す前に、Vertex AI クイックスタート: クライアント ライブラリの使用にある Node.js の設定手順を完了してください。詳細については、Vertex AI Node.js API のリファレンス ドキュメントをご覧ください。

Vertex AI に対する認証を行うには、アプリケーションのデフォルト認証情報を設定します。詳細については、ローカル開発環境の認証を設定するをご覧ください。

const {GoogleGenAI, Type} = require('@google/genai');

const GOOGLE_CLOUD_PROJECT = process.env.GOOGLE_CLOUD_PROJECT;
const GOOGLE_CLOUD_LOCATION = process.env.GOOGLE_CLOUD_LOCATION || 'global';
async function generateFunctionDesc(
  projectId = GOOGLE_CLOUD_PROJECT,
  location = GOOGLE_CLOUD_LOCATION
) {
  const client = new GoogleGenAI({
    vertexai: true,
    project: projectId,
    location: location,
  });

  const get_album_sales = {
    name: 'get_album_sales',
    description: 'Gets the number of albums sold',
    parameters: {
      type: Type.OBJECT,
      properties: {
        albums: {
          type: Type.ARRAY,
          description: 'List of albums',
          items: {
            description: 'Album and its sales',
            type: Type.OBJECT,
            properties: {
              album_name: {
                type: Type.STRING,
                description: 'Name of the music album',
              },
              copies_sold: {
                type: Type.INTEGER,
                description: 'Number of copies sold',
              },
            },
          },
        },
      },
    },
  };

  const sales_tool = {
    functionDeclarations: [get_album_sales],
  };

  const prompt = `
    At Stellar Sounds, a music label, 2024 was a rollercoaster. "Echoes of the Night", a debut synth-pop album, 
    surprisingly sold 350,000 copies, while veteran rock band "Crimson Tide's" latest, "Reckless Hearts",
    lagged at 120,000. Their up-and-coming indie artist, "Luna Bloom's" EP, "Whispers of Dawn",
    secured 75,000 sales. The biggest disappointment was the highly-anticipated rap album "Street Symphony" 
    only reaching 100,000 units. Overall, Stellar Sounds moved over 645,000 units this year, revealing unexpected
    trends in music consumption.
  `;

  const response = await client.models.generateContent({
    model: 'gemini-2.5-flash',
    contents: prompt,
    config: {
      tools: [sales_tool],
      temperature: 0,
    },
  });
  const output = JSON.stringify(response.functionCalls, null, 2);
  console.log(output);

  // Example response:
  //    [FunctionCall(
  //     id=None,
  //     name="get_album_sales",
  //     args={
  //         "albums": [
  //             {"album_name": "Echoes of the Night", "copies_sold": 350000},
  //             {"copies_sold": 120000, "album_name": "Reckless Hearts"},
  //             {"copies_sold": 75000, "album_name": "Whispers of Dawn"},
  //             {"copies_sold": 100000, "album_name": "Street Symphony"},
  //          ]
  //      },
  //     )]

  return output;
}

Python

このサンプルを試す前に、Vertex AI クイックスタート: クライアント ライブラリの使用にある Python の設定手順を完了してください。詳細については、Vertex AI Python API のリファレンス ドキュメントをご覧ください。

Vertex AI に対する認証を行うには、アプリケーションのデフォルト認証情報を設定します。詳細については、ローカル開発環境の認証を設定するをご覧ください。

from google import genai
from google.genai.types import (
    FunctionDeclaration,
    GenerateContentConfig,
    HttpOptions,
    Tool,
)

client = genai.Client(http_options=HttpOptions(api_version="v1"))
model_id = "gemini-2.5-flash"

get_album_sales = FunctionDeclaration(
    name="get_album_sales",
    description="Gets the number of albums sold",
    # Function parameters are specified in JSON schema format
    parameters={
        "type": "OBJECT",
        "properties": {
            "albums": {
                "type": "ARRAY",
                "description": "List of albums",
                "items": {
                    "description": "Album and its sales",
                    "type": "OBJECT",
                    "properties": {
                        "album_name": {
                            "type": "STRING",
                            "description": "Name of the music album",
                        },
                        "copies_sold": {
                            "type": "INTEGER",
                            "description": "Number of copies sold",
                        },
                    },
                },
            },
        },
    },
)

sales_tool = Tool(
    function_declarations=[get_album_sales],
)

response = client.models.generate_content(
    model=model_id,
    contents='At Stellar Sounds, a music label, 2024 was a rollercoaster. "Echoes of the Night," a debut synth-pop album, '
    'surprisingly sold 350,000 copies, while veteran rock band "Crimson Tide\'s" latest, "Reckless Hearts," '
    'lagged at 120,000. Their up-and-coming indie artist, "Luna Bloom\'s" EP, "Whispers of Dawn," '
    'secured 75,000 sales. The biggest disappointment was the highly-anticipated rap album "Street Symphony" '
    "only reaching 100,000 units. Overall, Stellar Sounds moved over 645,000 units this year, revealing unexpected "
    "trends in music consumption.",
    config=GenerateContentConfig(
        tools=[sales_tool],
        temperature=0,
    ),
)

print(response.function_calls)
# Example response:
# [FunctionCall(
#     id=None,
#     name="get_album_sales",
#     args={
#         "albums": [
#             {"album_name": "Echoes of the Night", "copies_sold": 350000},
#             {"copies_sold": 120000, "album_name": "Reckless Hearts"},
#             {"copies_sold": 75000, "album_name": "Whispers of Dawn"},
#             {"copies_sold": 100000, "album_name": "Street Symphony"},
#         ]
#     },
# )]

次のステップ

他の Google Cloud プロダクトのコードサンプルを検索およびフィルタするには、Google Cloud サンプル ブラウザをご覧ください。