Erste Schritte

In diesem Leitfaden erfahren Sie, wie Sie mit der Gemini API und der Interactions API beginnen. Sie führen Ihren ersten API-Aufruf in weniger als einer Minute aus und lernen die Textgenerierung, das multimodale Verständnis, die Bildgenerierung, die strukturierte Ausgabe, Tools, Funktionsaufrufe, Agents und die Hintergrundausführung kennen.

Die Interactions API ist über die Python- und JavaScript-SDKs sowie über REST verfügbar.

1. API-Schlüssel anfordern

Zur Verwendung der Gemini API benötigen Sie einen API-Schlüssel, um Ihre Anfragen zu authentifizieren, Sicherheitslimits durchzusetzen und die Nutzung für Ihr Konto zu erfassen.

  • In Google AI Studio werden für neue Nutzer automatisch ein Projekt und ein API-Schlüssel erstellt. Sie können ihn auf der Seite API-Schlüssel kopieren.
  • Wenn Sie einen neuen Schlüssel benötigen, klicken Sie in AI Studio auf API-Schlüssel erstellen und folgen Sie dem Dialogfeld, um ein neues Schlüssel-Projekt-Paar hinzuzufügen.

Gemini API-Schlüssel erstellen

Legen Sie Ihren Schlüssel als Umgebungsvariable fest:

export GEMINI_API_KEY="YOUR_API_KEY"

Upgrade auf die kostenpflichtige Stufe durchführen

Wenn Sie auf die kostenpflichtige Stufe upgraden, erhöhen sich Ihre Ratenbegrenzungen. Außerdem müssen Sie Cloud Billing einrichten.

  • Klicken Sie auf den Seiten API-Schlüssel oder Projekte von AI Studio auf Abrechnung einrichten.
  • Folgen Sie dem Cloud Billing-Dialogfeld, um ein Rechnungskonto zu erstellen oder zu verknüpfen, eine Zahlungsmethode hinzuzufügen und mindestens 5 $ (oder den entsprechenden Betrag in Ihrer Landeswährung) als Guthaben zu bezahlen.
  • Ihre API-Nutzung können Sie in Google AI Studio unter Dashboard > Nutzung einsehen.

Weitere Informationen finden Sie auf der Abrechnungsseite.

2. SDK installieren und ersten Aufruf ausführen

Installieren Sie das SDK und generieren Sie Text mit einem einzigen API-Aufruf.

Python

Installieren Sie das SDK:

pip install -U google-genai

Client initialisieren und Anfrage stellen:

from google import genai

client = genai.Client()

interaction = client.interactions.create(
    model="gemini-3.8-flash",
    input="Explain how AI works in a few words"
)
print(interaction.output_text)

JavaScript

Installieren Sie das SDK:

npm install @google/genai

Client initialisieren und Anfrage stellen:

import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({});

const interaction = await ai.interactions.create({
  model: "gemini-3.8-flash",
  input: "Explain how AI works in a few words",
});
console.log(interaction.output_text);

Java

import com.google.genai.Client;
import com.google.genai.gaos.models.interactions.CreateModelInteraction;
import com.google.genai.gaos.models.interactions.Interaction;
import com.google.genai.gaos.models.interactions.InteractionsInput;
import com.google.genai.gaos.models.interactions.Model;
import com.google.genai.gaos.models.operations.CreateInteractionRequestBody;

Client client = new Client();

CreateModelInteraction params =
    CreateModelInteraction.builder()
        .model(Model.of("gemini-3.8-flash"))
        .input(InteractionsInput.of("Explain how AI works in a few words"))
        .build();

Interaction interaction =
    client.interactions.create(CreateInteractionRequestBody.of(params)).interaction().get();

System.out.println(interaction.outputText().orElse(""));

Go

package main

import (
    "context"
    "fmt"
    "log"

    "google.golang.org/genai"
    "google.golang.org/genai/interactions/models/interactions"
    "google.golang.org/genai/interactions/models/operations"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    res, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model: interactions.Model("gemini-3.8-flash"),
            Input: interactions.NewInteractionsInput("Explain how AI works in a few words."),
        }),
    })
    if err != nil {
        log.Fatal(err)
    }

    if res.Interaction.OutputText != nil {
        fmt.Println(*res.Interaction.OutputText)
    }
}

REST

curl -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemini-3.8-flash",
    "input": "Explain how AI works in a few words"
  }'

Antwort

{
  "id": "v1_ChdpQUFvYXI...",
  "status": "completed",
  "usage": {
    "total_tokens": 197,
    "total_input_tokens": 8,
    "total_output_tokens": 12
  },
  "created": "2026-06-09T12:01:25Z",
  "steps": [
    {
      "type": "thought",
      "signature": "EvEFCu4FAQw..."
    },
    {
      "type": "model_output",
      "content": [
        {
          "type": "text",
          "text": "AI learns patterns from data, then uses those patterns to make predictions or decisions on new data."
        }
      ]
    }
  ],
  "object": "interaction",
  "model": "gemini-3.8-flash",
}

Bei Verwendung von REST gibt die API die vollständige Interaction-Ressource mit Metadaten, Nutzungsstatistiken und dem detaillierten Verlauf des Zuges zurück.

Die SDKs machen zwar die vollständige Antwort verfügbar, bieten aber auch praktische Eigenschaften wie interaction.output_text und interaction.output_image, um direkt auf die endgültigen Ausgaben zuzugreifen. Weitere Informationen zur Antwortstruktur finden Sie in der Übersicht zu Interaktionen. Details zu Systemanweisungen und der Generierungskonfiguration finden Sie im Leitfaden zur Textgenerierung.

3. Antwort streamen

Um flüssigere Interaktionen zu ermöglichen, können Sie die Antwort streamen, während sie generiert wird. Bei jedem step.delta-Ereignis wird ein Textblock zurückgegeben, den Sie sofort anzeigen können.

Python

from google import genai

client = genai.Client()

stream = client.interactions.create(
    model="gemini-3.8-flash",
    input="Explain how AI works",
    stream=True
)
for event in stream:
    print(event)

JavaScript

import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({});

const stream = await ai.interactions.create({
  model: "gemini-3.8-flash",
  input: "Explain how AI works",
  stream: true,
});

for await (const event of stream) {
  console.log(event);
}

Java

import com.google.genai.Client;
import com.google.genai.gaos.models.interactions.CreateModelInteraction;
import com.google.genai.gaos.models.interactions.InteractionSSEStreamEvent;
import com.google.genai.gaos.models.interactions.InteractionsInput;
import com.google.genai.gaos.models.interactions.Model;
import com.google.genai.gaos.models.operations.CreateInteractionRequestBody;
import com.google.genai.gaos.models.operations.CreateInteractionResponse;
import com.google.genai.gaos.utils.EventStream;

Client client = new Client();

CreateModelInteraction params =
    CreateModelInteraction.builder()
        .model(Model.of("gemini-3.8-flash"))
        .input(InteractionsInput.of("Explain how AI works"))
        .stream(true)
        .build();

CreateInteractionResponse response =
    client.interactions.create(CreateInteractionRequestBody.of(params));

try (EventStream<InteractionSSEStreamEvent> stream = response.events()) {
  for (InteractionSSEStreamEvent event : stream) {
    System.out.println(event);
  }
}

Go

package main

import (
    "context"
    "fmt"
    "log"

    "google.golang.org/genai"
    "google.golang.org/genai/interactions/models/interactions"
    "google.golang.org/genai/interactions/models/operations"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    res, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model: interactions.Model("gemini-3.8-flash"),
            Input: interactions.NewInteractionsInput("Write a haiku about coding."),
        }),
    })
    if err != nil {
        log.Fatal(err)
    }

    if res.Interaction.OutputText != nil {
        fmt.Println(*res.Interaction.OutputText)
    }
}

REST

curl -X POST "https://generativelanguage.googleapis.com/v1beta/interactions?alt=sse" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  --no-buffer \
  -d '{
    "model": "gemini-3.8-flash",
    "input": "Explain how AI works",
    "stream": true
  }'

Beim Streaming antwortet der Server mit einem Stream von Server-Sent Events (SSE). Jedes Ereignis enthält einen Typ und JSON-Daten.

Antwort

event: interaction.created
data: {"interaction":{"id":"v1_Chd...","status":"in_progress","model":"gemini-3.8-flash"},"event_type":"interaction.created"}

event: step.start
data: {"index":0,"step":{"type":"thought"},"event_type":"step.start"}

event: step.delta
data: {"index":0,"delta":{"signature":"EvEFCu4F...","type":"thought_signature"},"event_type":"step.delta"}

event: step.stop
data: {"index":0,"event_type":"step.stop"}

event: step.start
data: {"index":1,"step":{"type":"model_output"},"event_type":"step.start"}

event: step.delta
data: {"index":1,"delta":{"text":"AI ","type":"text"},"event_type":"step.delta"}

event: step.delta
data: {"index":1,"delta":{"text":"works ","type":"text"},"event_type":"step.delta"}

event: step.stop
data: {"index":1,"event_type":"step.stop"}

event: interaction.completed
data: {"interaction":{"id":"v1_Chd...","status":"completed","usage":{"total_tokens":197}},"event_type":"interaction.completed"}

Eine detaillierte Beschreibung der Verarbeitung von Streaming-Ereignissen und ‑Deltatypen finden Sie im Leitfaden zu Streaming-Interaktionen.

4. Unterhaltungen über mehrere Themen

Die Interactions API unterstützt Mehrfachdialog-Unterhaltungen mit zwei Ansätzen:

  • Zustandsbehaftet (empfohlen): Setzen Sie eine Unterhaltung auf dem Server mit previous_interaction_id fort. Ideal für die meisten Chat- und agentischen Workflows, bei denen der Server den Verlauf verwalten und das Caching optimieren soll.

  • Zustandslos: Der Gesprächsverlauf wird auf dem Client verwaltet, indem alle vorherigen Züge (einschließlich der Zwischenschritte für das Modell und das Tool) in jeder Anfrage übergeben werden.

Ketten Sie Interaktionen, indem Sie previous_interaction_id übergeben. Der Server verwaltet den gesamten Unterhaltungsverlauf für Sie.

Python

from google import genai

client = genai.Client()

# Server-side state (recommended)
interaction1 = client.interactions.create(
    model="gemini-3.8-flash",
    input="I have 2 dogs in my house.",
)
print("Response 1:", interaction1.output_text)

interaction2 = client.interactions.create(
    model="gemini-3.8-flash",
    input="How many paws are in my house?",
    previous_interaction_id=interaction1.id,
)
print("Response 2:", interaction2.output_text)

JavaScript

import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({});

// Server-side state (recommended)
const interaction1 = await ai.interactions.create({
  model: "gemini-3.8-flash",
  input: "I have 2 dogs in my house.",
});
console.log("Response 1:", interaction1.output_text);

const interaction2 = await ai.interactions.create({
  model: "gemini-3.8-flash",
  input: "How many paws are in my house?",
  previous_interaction_id: interaction1.id,
});
console.log("Response 2:", interaction2.output_text);

Java

import com.google.genai.Client;
import com.google.genai.gaos.models.interactions.CreateModelInteraction;
import com.google.genai.gaos.models.interactions.Interaction;
import com.google.genai.gaos.models.interactions.InteractionsInput;
import com.google.genai.gaos.models.interactions.Model;
import com.google.genai.gaos.models.operations.CreateInteractionRequestBody;

Client client = new Client();

// Server-side state (recommended)
CreateModelInteraction params1 =
    CreateModelInteraction.builder()
        .model(Model.of("gemini-3.8-flash"))
        .input(InteractionsInput.of("I have 2 dogs in my house."))
        .build();

Interaction interaction1 =
    client.interactions.create(CreateInteractionRequestBody.of(params1)).interaction().get();
System.out.println("Response 1: " + interaction1.outputText().orElse(""));

CreateModelInteraction params2 =
    CreateModelInteraction.builder()
        .model(Model.of("gemini-3.8-flash"))
        .input(InteractionsInput.of("How many paws are in my house?"))
        .previousInteractionId(interaction1.id().orElse(""))
        .build();

Interaction interaction2 =
    client.interactions.create(CreateInteractionRequestBody.of(params2)).interaction().get();
System.out.println("Response 2: " + interaction2.outputText().orElse(""));

Go

package main

import (
    "context"
    "fmt"
    "log"

    "google.golang.org/genai"
    "google.golang.org/genai/interactions/models/interactions"
    "google.golang.org/genai/interactions/models/operations"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    res, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model:  interactions.Model("gemini-3.8-flash"),
            Input:  interactions.NewInteractionsInput("Explain quantum computing in simple terms."),
            Stream: genai.Ptr(true),
        }),
    })
    if err != nil {
        log.Fatal(err)
    }
    stream := res.InteractionSSEStreamEvent
    defer stream.Close()

    for stream.Next() {
        event := stream.Value()
        if stepDelta := event.GetDataStepDelta(); stepDelta != nil {
            if textDelta := stepDelta.GetDeltaText(); textDelta != nil {
                fmt.Print(textDelta.GetText())
            }
        }
    }
    if err := stream.Err(); err != nil {
        log.Fatal(err)
    }
}

REST

RESPONSE1=$(curl -s -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemini-3.8-flash",
    "input": "I have 2 dogs in my house."
  }')

INTERACTION_ID=$(echo "$RESPONSE1" | jq -r '.id')
echo "Interaction 1 ID: $INTERACTION_ID"

curl -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemini-3.8-flash",
    "input": "How many paws are in my house?",
    "previous_interaction_id": "'$INTERACTION_ID'"
  }'

Zustandslos

store=false festlegen und den Unterhaltungsverlauf auf der Clientseite verwalten. Sie müssen alle vom Modell generierten Schritte (einschließlich der Schritte thought und function_call) genau wie erhalten beibehalten und noch einmal senden.

Python

from google import genai

client = genai.Client()

history = [
    {
        "type": "user_input",
        "content": [{"type": "text", "text": "I have 2 dogs in my house."}]
    }
]

interaction1 = client.interactions.create(
    model="gemini-3.8-flash",
    store=False,
    input=history
)
print("Response 1:", interaction1.steps[-1].content[0].text)

for step in interaction1.steps:
    history.append(step.model_dump())

history.append({
    "type": "user_input",
    "content": [{"type": "text", "text": "How many paws are in my house?"}]
})

interaction2 = client.interactions.create(
    model="gemini-3.8-flash",
    store=False,
    input=history
)
print("Response 2:", interaction2.steps[-1].content[0].text)

JavaScript

import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({});

const history = [
  {
    type: "user_input",
    content: [{ type: "text", text: "I have 2 dogs in my house." }]
  }
];

const interaction1 = await ai.interactions.create({
  model: "gemini-3.8-flash",
  store: false,
  input: history
});
console.log("Response 1:", interaction1.steps.at(-1).content[0].text);

history.push(...interaction1.steps);

history.push({
  type: "user_input",
  content: [{ type: "text", text: "How many paws are in my house?" }]
});

const interaction2 = await ai.interactions.create({
  model: "gemini-3.8-flash",
  store: false,
  input: history
});
console.log("Response 2:", interaction2.steps.at(-1).content[0].text);

Java

import com.google.genai.Client;
import com.google.genai.gaos.models.interactions.CreateModelInteraction;
import com.google.genai.gaos.models.interactions.Interaction;
import com.google.genai.gaos.models.interactions.InteractionsInput;
import com.google.genai.gaos.models.interactions.Model;
import com.google.genai.gaos.models.interactions.Step;
import com.google.genai.gaos.models.interactions.TextContent;
import com.google.genai.gaos.models.interactions.UserInputStep;
import com.google.genai.gaos.models.operations.CreateInteractionRequestBody;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.List;

Client client = new Client();

List<Step> history = new ArrayList<>();
history.add(
    UserInputStep.builder()
        .content(Arrays.asList(TextContent.builder().text("I have 2 dogs in my house.").build()))
        .build());

CreateModelInteraction params1 =
    CreateModelInteraction.builder()
        .model(Model.of("gemini-3.8-flash"))
        .store(false)
        .input(InteractionsInput.ofStep(history))
        .build();

Interaction interaction1 =
    client.interactions.create(CreateInteractionRequestBody.of(params1)).interaction().get();
System.out.println("Response 1: " + interaction1.outputText().orElse(""));

interaction1.steps().ifPresent(history::addAll);

history.add(
    UserInputStep.builder()
        .content(Arrays.asList(TextContent.builder().text("How many paws are in my house?").build()))
        .build());

CreateModelInteraction params2 =
    CreateModelInteraction.builder()
        .model(Model.of("gemini-3.8-flash"))
        .store(false)
        .input(InteractionsInput.ofStep(history))
        .build();

Interaction interaction2 =
    client.interactions.create(CreateInteractionRequestBody.of(params2)).interaction().get();
System.out.println("Response 2: " + interaction2.outputText().orElse(""));

Go

package main

import (
    "context"
    "fmt"
    "log"

    "google.golang.org/genai"
    "google.golang.org/genai/interactions/models/interactions"
    "google.golang.org/genai/interactions/models/operations"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    // 1. First turn
    res1, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model: interactions.Model("gemini-3.8-flash"),
            Input: interactions.NewInteractionsInput("Hi, my name is Alex."),
        }),
    })
    if err != nil {
        log.Fatal(err)
    }
    if res1.Interaction.OutputText != nil {
        fmt.Println(*res1.Interaction.OutputText)
    }

    // 2. Second turn (passing PreviousInteractionID)
    res2, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model:                 interactions.Model("gemini-3.8-flash"),
            Input:                 interactions.NewInteractionsInput("What's my name?"),
            PreviousInteractionID: res1.Interaction.ID,
        }),
    })
    if err != nil {
        log.Fatal(err)
    }
    if res2.Interaction.OutputText != nil {
        fmt.Println(*res2.Interaction.OutputText)
    }
}

REST

# Turn 1: Send with store: false
RESPONSE1=$(curl -s -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemini-3.8-flash",
    "store": false,
    "input": [
      {
        "type": "user_input",
        "content": "I have 2 dogs in my house."
      }
    ]
  }')

MODEL_STEPS=$(echo "$RESPONSE1" | jq '.steps')

# Turn 2: Build full history
HISTORY=$(jq -n \
  --argjson first_input '[{"type": "user_input", "content": "I have 2 dogs in my house."}]' \
  --argjson model_steps "$MODEL_STEPS" \
  --argjson second_input '[{"type": "user_input", "content": "How many paws are in my house?"}]' \
  '$first_input + $model_steps + $second_input')

curl -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d "{
    \"model\": \"gemini-3.8-flash\",
    \"store\": false,
    \"input\": $HISTORY
  }"

Antwort

{
  "id": "v2_Chd...",
  "status": "completed",
  "usage": {
    "total_tokens": 240,
    "total_input_tokens": 60,
    "total_output_tokens": 20
  },
  "steps": [
    {
      "type": "model_output",
      "content": [
        {
          "type": "text",
          "text": "There are 8 paws in your house. 2 dogs \u00d7 4 paws = 8 paws."
        }
      ]
    }
  ],
  "object": "interaction",
  "model": "gemini-3.8-flash"
}

Bei der zweiten Interaktion wird ein vollständiges Antwortobjekt zurückgegeben, das nur die neuen Schritte enthält, aber auf dem Kontext des vorherigen Turns basiert. Weitere Informationen zum Beibehalten des Zustands in Mehrfachdialogen oder zum zustandslosen Modus für die clientseitige Verlaufsverwaltung

5. Multimodales Verstehen

Gemini-Modelle können Bilder, Audio, Video und Dokumente nativ verstehen. Media zusammen mit Text in einer einzelnen Anfrage übergeben.

Python

import base64
from google import genai

client = genai.Client()

# Load a local image
with open("sample.jpg", "rb") as f:
    image_bytes = f.read()
image_b64 = base64.b64encode(image_bytes).decode("utf-8")

interaction = client.interactions.create(
    model="gemini-3.8-flash",
    input=[
        {"type": "text", "text": "Compare this local image and this remote audio file."},
        {
            "type": "image",
            "data": image_b64,
            "mime_type": "image/jpeg"
        },
        {
            "type": "audio",
            "uri": "https://storage.googleapis.com/generativeai-downloads/data/sample.mp3",
            "mime_type": "audio/mp3"
        }
    ]
)
print(interaction.output_text)

JavaScript

import fs from "fs";
import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({});

// Load a local image
const imageBytes = fs.readFileSync("sample.jpg");
const imageB64 = imageBytes.toString("base64");

const interaction = await ai.interactions.create({
  model: "gemini-3.8-flash",
  input: [
    { type: "text", text: "Compare this local image and this remote audio file." },
    {
      type: "image",
      data: imageB64,
      mime_type: "image/jpeg"
    },
    {
      type: "audio",
      uri: "https://storage.googleapis.com/generativeai-downloads/data/sample.mp3",
      mime_type: "audio/mp3"
    }
  ],
});
console.log(interaction.output_text);

Java

import com.google.genai.Client;
import com.google.genai.gaos.models.interactions.AudioContent;
import com.google.genai.gaos.models.interactions.AudioContentMimeType;
import com.google.genai.gaos.models.interactions.CreateModelInteraction;
import com.google.genai.gaos.models.interactions.ImageContent;
import com.google.genai.gaos.models.interactions.ImageContentMimeType;
import com.google.genai.gaos.models.interactions.Interaction;
import com.google.genai.gaos.models.interactions.InteractionsInput;
import com.google.genai.gaos.models.interactions.Model;
import com.google.genai.gaos.models.interactions.TextContent;
import com.google.genai.gaos.models.operations.CreateInteractionRequestBody;
import java.nio.file.Files;
import java.nio.file.Path;
import java.util.Arrays;
import java.util.Base64;

Client client = new Client();

// Load a local image
byte[] imageBytes = Files.readAllBytes(Path.of("sample.jpg"));
String imageB64 = Base64.getEncoder().encodeToString(imageBytes);

CreateModelInteraction params =
    CreateModelInteraction.builder()
        .model(Model.of("gemini-3.8-flash"))
        .input(
            InteractionsInput.ofContent(
                Arrays.asList(
                    TextContent.builder()
                        .text("Compare this local image and this remote audio file.")
                        .build(),
                    ImageContent.builder()
                        .data(imageB64)
                        .mimeType(ImageContentMimeType.IMAGE_JPEG)
                        .build(),
                    AudioContent.builder()
                        .uri("https://storage.googleapis.com/generativeai-downloads/data/sample.mp3")
                        .mimeType(AudioContentMimeType.AUDIO_MP3)
                        .build())))
        .build();

Interaction interaction =
    client.interactions.create(CreateInteractionRequestBody.of(params)).interaction().get();
System.out.println(interaction.outputText().orElse(""));

Go

package main

import (
    "context"
    "encoding/base64"
    "fmt"
    "log"
    "os"

    "google.golang.org/genai"
    "google.golang.org/genai/interactions/models/interactions"
    "google.golang.org/genai/interactions/models/operations"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    imageBytes, err := os.ReadFile("path/to/organ.jpg")
    if err != nil {
        log.Fatal(err)
    }
    base64Image := base64.StdEncoding.EncodeToString(imageBytes)

    res, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model: interactions.Model("gemini-3.8-flash"),
            Input: interactions.NewInteractionsInput([]interactions.Content{
                interactions.NewContent(interactions.TextContent{
                    Text: "What is in this image?",
                }),
                interactions.NewContent(interactions.ImageContent{
                    Data:     genai.Ptr(base64Image),
                    MimeType: interactions.ImageContentMimeTypeImageJpeg.ToPointer(),
                }),
            }),
        }),
    })
    if err != nil {
        log.Fatal(err)
    }

    if res.Interaction.OutputText != nil {
        fmt.Println(*res.Interaction.OutputText)
    }
}

REST

# Base64-encode local image
BASE64_IMAGE=$(base64 -w 0 sample.jpg)

curl -X POST "https://generativelanguage.googleapis.com/v1beta/interactions"   -H "x-goog-api-key: $GEMINI_API_KEY"   -H 'Content-Type: application/json'   -H "Api-Revision: 2026-05-20"   -d '{
    "model": "gemini-3.8-flash",
    "input": [
      {
        "type": "text",
        "text": "Compare this local image and this remote audio file."
      },
      {
        "type": "image",
        "data": "'$BASE64_IMAGE'",
        "mime_type": "image/jpeg"
      },
      {
        "type": "audio",
        "uri": "https://storage.googleapis.com/generativeai-downloads/data/sample.mp3",
        "mime_type": "audio/mp3"
      }
    ]
  }'

Antwort

{
  "id": "v1_Chd...",
  "status": "completed",
  "usage": {
    "total_tokens": 300
  },
  "steps": [
    {
      "type": "model_output",
      "content": [
        {
          "type": "text",
          "text": "The local image displays a pipe organ while the remote audio file is a sample MP3 clip..."
        }
      ]
    }
  ],
  "object": "interaction",
  "model": "gemini-3.8-flash",
}

Informationen zum Übergeben von Bildern, Videos und Audiodateien finden Sie im Leitfaden zum Bildverständnis.

Verständnis von Audioinhalten

Audioinhalte transkribieren, zusammenfassen oder Fragen dazu beantworten lassen

Videos verstehen

Videoinhalte analysieren, Ereignisse lokalisieren und Aktionen beschreiben

Dokumentverarbeitung

Informationen aus PDFs und anderen Dokumentformaten extrahieren

6. Multimodale Generierung

Gemini kann Bilder nativ mit den Nano Banana-Bildmodellen generieren.

Python

import base64
from google import genai

client = genai.Client()

interaction = client.interactions.create(
    model="gemini-3.1-flash-image",
    input="Generate an image of a futuristic city skyline at sunset",
)

with open("generated_image.png", "wb") as f:
    f.write(base64.b64decode(interaction.output_image.data))

JavaScript

import { GoogleGenAI } from "@google/genai";
import * as fs from "node:fs";

const ai = new GoogleGenAI({});

const interaction = await ai.interactions.create({
  model: "gemini-3.1-flash-image",
  input: "Generate an image of a futuristic city skyline at sunset",
});

const generatedImage = interaction.output_image;
if (generatedImage) {
  const buffer = Buffer.from(generatedImage.data, "base64");
  fs.writeFileSync("generated_image.png", buffer);
}

Java

import com.google.genai.Client;
import com.google.genai.gaos.models.interactions.CreateModelInteraction;
import com.google.genai.gaos.models.interactions.ImageContent;
import com.google.genai.gaos.models.interactions.Interaction;
import com.google.genai.gaos.models.interactions.InteractionsInput;
import com.google.genai.gaos.models.interactions.Model;
import com.google.genai.gaos.models.operations.CreateInteractionRequestBody;
import java.nio.file.Files;
import java.nio.file.Path;
import java.util.Base64;

Client client = new Client();

CreateModelInteraction params =
    CreateModelInteraction.builder()
        .model(Model.of("gemini-3.1-flash-image"))
        .input(InteractionsInput.of("Generate an image of a futuristic city skyline at sunset"))
        .build();

Interaction interaction =
    client.interactions.create(CreateInteractionRequestBody.of(params)).interaction().get();

if (interaction.outputImage().isPresent()) {
  ImageContent generatedImage = interaction.outputImage().get();
  if (generatedImage.data().isPresent()) {
    byte[] imageBytes = Base64.getDecoder().decode(generatedImage.data().get());
    Files.write(Path.of("generated_image.png"), imageBytes);
  }
}

Go

package main

import (
    "context"
    "fmt"
    "log"

    "google.golang.org/genai"
    "google.golang.org/genai/interactions/models/interactions"
    "google.golang.org/genai/interactions/models/operations"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    weatherTool := interactions.NewTool(interactions.Function{
        Name:        genai.Ptr("get_current_weather"),
        Description: genai.Ptr("Gets the current weather for a given location."),
        Parameters: map[string]any{
            "type": "object",
            "properties": map[string]any{
                "location": map[string]any{
                    "type":        "string",
                    "description": "The city and state, e.g. San Francisco, CA",
                },
            },
            "required": []string{"location"},
        },
    })

    // 1. Send prompt with tool declaration
    res, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model: interactions.Model("gemini-3.8-flash"),
            Input: interactions.NewInteractionsInput("What is the weather like in Boston?"),
            Tools: []interactions.Tool{weatherTool},
        }),
    })
    if err != nil {
        log.Fatal(err)
    }

    // 2. Check if the model requested a function call
    for _, step := range res.Interaction.Steps {
        if call := step.FunctionCallStep; call != nil {
            fmt.Printf("Function to call: %s\n", call.Name)
            fmt.Printf("Arguments: %v\n", call.Arguments)

            // 3. Execute your local function and send the result back
            finalRes, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
                Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
                    Model:                 interactions.Model("gemini-3.8-flash"),
                    PreviousInteractionID: res.Interaction.ID,
                    Input: interactions.NewInteractionsInput([]interactions.Step{
                        interactions.NewStep(interactions.FunctionResultStep{
                            Name:   genai.Ptr(call.Name),
                            CallID: call.ID,
                            Result: interactions.NewFunctionResultStepResultUnion(`{"temperature": "72F", "condition": "Sunny"}`),
                        }),
                    }),
                }),
            })
            if err != nil {
                log.Fatal(err)
            }

            if finalRes.Interaction.OutputText != nil {
                fmt.Println(*finalRes.Interaction.OutputText)
            }
        }
    }
}

REST

curl -s -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemini-3.1-flash-image",
    "input": [
      {"type": "text", "text": "Generate an image of a futuristic city skyline at sunset"}
    ]
  }'

Antwort

{
  "id": "v1_Chd...",
  "status": "completed",
  "steps": [
    {
      "type": "model_output",
      "content": [
        {
          "type": "image",
          "data": "BASE64_ENCODED_IMAGE",
          "mime_type": "image/png"
        }
      ]
    }
  ],
  "object": "interaction",
  "model": "gemini-3.1-flash-image",
}

Wenn das Modell ein Bild generiert, gibt es die base64-codierten Bilddaten in einem Schritt innerhalb des steps-Arrays sowie über die Convenience-Property output_image zurück. Im Leitfaden zur Bildgenerierung finden Sie Informationen zu Seitenverhältnissen, Bildbearbeitung und Referenzen.

Sprachgenerierung

Mit Gemini 3.1 Flash TTS ausdrucksstarke Sprache mit mehreren Sprechern generieren

Musikgenerierung

Mit Lyria 3.5 kannst du Clips und Songs in voller Länge erstellen.

7. Strukturierte Ausgabe verwenden

Konfigurieren Sie das Modell so, dass es JSON zurückgibt, das einem von Ihnen definierten Schema entspricht. Die strukturierte Ausgabe funktioniert mit Pydantic (Python) und Zod (JavaScript).

Python

from google import genai
from pydantic import BaseModel, Field
from typing import List, Optional

class Recipe(BaseModel):
    recipe_name: str = Field(description="Name of the recipe.")
    ingredients: List[str] = Field(description="List of ingredients.")
    prep_time_minutes: Optional[int] = Field(description="Prep time in minutes.")

client = genai.Client()

interaction = client.interactions.create(
    model="gemini-3.8-flash",
    input="Give me a recipe for banana bread",
    response_format={
        "type": "text",
        "mime_type": "application/json",
        "schema": Recipe.model_json_schema()
    },
)

recipe = Recipe.model_validate_json(interaction.output_text)
print(recipe)

JavaScript

import { GoogleGenAI } from "@google/genai";
import * as z from "zod";

const ai = new GoogleGenAI({});

const recipeJsonSchema = {
  type: "object",
  properties: {
    recipe_name: { type: "string", description: "Name of the recipe." },
    ingredients: {
      type: "array",
      items: { type: "string" },
      description: "List of ingredients."
    },
    prep_time_minutes: {
      type: "integer",
      description: "Prep time in minutes."
    }
  },
  required: ["recipe_name", "ingredients"]
};

const recipeSchema = z.fromJSONSchema(recipeJsonSchema);

const interaction = await ai.interactions.create({
  model: "gemini-3.8-flash",
  input: "Give me a recipe for banana bread",
  response_format: {
    type: "text",
    mime_type: "application/json",
    schema: recipeJsonSchema
  },
});

const recipe = recipeSchema.parse(JSON.parse(interaction.output_text));
console.log(recipe);

Java

import com.google.genai.Client;
import com.google.genai.gaos.models.interactions.CreateModelInteraction;
import com.google.genai.gaos.models.interactions.CreateModelInteractionResponseFormat;
import com.google.genai.gaos.models.interactions.Interaction;
import com.google.genai.gaos.models.interactions.InteractionsInput;
import com.google.genai.gaos.models.interactions.Model;
import com.google.genai.gaos.models.interactions.ResponseFormat;
import com.google.genai.gaos.models.interactions.TextResponseFormat;
import com.google.genai.gaos.models.interactions.TextResponseFormatMimeType;
import com.google.genai.gaos.models.operations.CreateInteractionRequestBody;
import java.util.Arrays;
import java.util.HashMap;
import java.util.Map;

Client client = new Client();

Map<String, Object> recipeNameProp = new HashMap<>();
recipeNameProp.put("type", "string");
recipeNameProp.put("description", "Name of the recipe.");

Map<String, Object> itemsProp = new HashMap<>();
itemsProp.put("type", "string");

Map<String, Object> ingredientsProp = new HashMap<>();
ingredientsProp.put("type", "array");
ingredientsProp.put("items", itemsProp);
ingredientsProp.put("description", "List of ingredients.");

Map<String, Object> prepTimeProp = new HashMap<>();
prepTimeProp.put("type", "integer");
prepTimeProp.put("description", "Prep time in minutes.");

Map<String, Object> properties = new HashMap<>();
properties.put("recipe_name", recipeNameProp);
properties.put("ingredients", ingredientsProp);
properties.put("prep_time_minutes", prepTimeProp);

Map<String, Object> recipeJsonSchema = new HashMap<>();
recipeJsonSchema.put("type", "object");
recipeJsonSchema.put("properties", properties);
recipeJsonSchema.put("required", Arrays.asList("recipe_name", "ingredients"));

CreateModelInteractionResponseFormat format =
    CreateModelInteractionResponseFormat.of(
        ResponseFormat.of(
            TextResponseFormat.builder()
                .mimeType(TextResponseFormatMimeType.APPLICATION_JSON)
                .schema(recipeJsonSchema)
                .build()));

CreateModelInteraction params =
    CreateModelInteraction.builder()
        .model(Model.of("gemini-3.8-flash"))
        .input(InteractionsInput.of("Give me a recipe for banana bread"))
        .responseFormat(format)
        .build();

Interaction interaction =
    client.interactions.create(CreateInteractionRequestBody.of(params)).interaction().get();
System.out.println(interaction.outputText().orElse(""));

Go

package main

import (
    "context"
    "fmt"
    "log"

    "google.golang.org/genai"
    "google.golang.org/genai/interactions/models/interactions"
    "google.golang.org/genai/interactions/models/operations"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    res, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model: interactions.Model("gemini-3.8-flash"),
            Input: interactions.NewInteractionsInput("Who won the latest Super Bowl and what was the score?"),
            Tools: []interactions.Tool{
                interactions.NewTool(interactions.GoogleSearch{}),
            },
        }),
    })
    if err != nil {
        log.Fatal(err)
    }

    if res.Interaction.OutputText != nil {
        fmt.Println(*res.Interaction.OutputText)
    }

    // Optional: Inspect search queries and citations
    for _, step := range res.Interaction.Steps {
        if searchCall := step.GoogleSearchCallStep; searchCall != nil {
            fmt.Printf("Search queries: %v\n", searchCall.Arguments.Queries)
        } else if modelOut := step.ModelOutputStep; modelOut != nil {
            for _, part := range modelOut.Content {
                if textPart := part.TextContent; textPart != nil {
                    for _, annotation := range textPart.Annotations {
                        if citation := annotation.URLCitation; citation != nil {
                            var title, url string
                            if citation.Title != nil {
                                title = *citation.Title
                            }
                            if citation.URL != nil {
                                url = *citation.URL
                            }
                            fmt.Printf("Source: %s (%s)\n", title, url)
                        }
                    }
                }
            }
        }
    }
}

REST

curl -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemini-3.8-flash",
    "input": "Give me a recipe for banana bread",
    "response_format": {
      "type": "text",
      "mime_type": "application/json",
      "schema": {
        "type": "object",
        "properties": {
          "recipe_name": { "type": "string", "description": "Name of the recipe." },
          "ingredients": {
            "type": "array",
            "items": { "type": "string" },
            "description": "List of ingredients."
          },
          "prep_time_minutes": {
            "type": "integer",
            "description": "Prep time in minutes."
          }
        },
        "required": ["recipe_name", "ingredients"]
      }
    }
  }'

Antwort

{
  "id": "v1_Chd...",
  "status": "completed",
  "steps": [
    {
      "type": "model_output",
      "content": [
        {
          "type": "text",
          "text": "{\n  \"recipe_name\": \"Classic Banana Bread\",\n  \"ingredients\": [\n    \"3 ripe bananas, mashed\",\n    \"1/3 cup melted butter\",\n    \"3/4 cup sugar\",\n    \"1 egg, beaten\",\n    \"1 teaspoon vanilla extract\",\n    \"1 teaspoon baking soda\",\n    \"Pinch of salt\",\n    \"1.5 cups all-purpose flour\"\n  ],\n  \"prep_time_minutes\": 15\n}"
        }
      ]
    }
  ],
  "object": "interaction",
  "model": "gemini-3.8-flash",
}

Der Ausgabetextblock enthält einen gültigen JSON-String, der genau dem angeforderten Schema entspricht. Informationen zum Definieren komplexerer Strukturen und rekursiver Schemas finden Sie im Leitfaden zur strukturierten Ausgabe.

8. Tools verwenden

Die Antwort des Modells mit Echtzeitinformationen aus der Google Suche fundieren. Die API führt automatisch eine Suche durch, verarbeitet die Ergebnisse und gibt Zitationen zurück.

Python

from google import genai

client = genai.Client()

interaction = client.interactions.create(
    model="gemini-3.8-flash",
    input="Who won the euro 2024?",
    tools=[{"type": "google_search"}]
)

print(interaction.output_text)

# Print citations
for step in interaction.steps:
    if step.type == "model_output":
        for content_block in step.content:
            if content_block.type == "text" and content_block.annotations:
                print("\nCitations:")
                for annotation in content_block.annotations:
                    if annotation.type == "url_citation":
                        print(f"  [{annotation.title}]({annotation.url})")

JavaScript

import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({});

const interaction = await ai.interactions.create({
  model: "gemini-3.8-flash",
  input: "Who won the euro 2024?",
  tools: [{ type: "google_search" }]
});

console.log(interaction.output_text);

// Print citations
for (const step of interaction.steps) {
  if (step.type === "model_output") {
    for (const contentBlock of step.content) {
      if (contentBlock.type === "text" && contentBlock.annotations) {
        console.log("\nCitations:");
        for (const annotation of contentBlock.annotations) {
          if (annotation.type === "url_citation") {
            console.log(`  [${annotation.title}](${annotation.url})`);
          }
        }
      }
    }
  }
}

Java

import com.google.genai.Client;
import com.google.genai.gaos.models.interactions.Annotation;
import com.google.genai.gaos.models.interactions.Content;
import com.google.genai.gaos.models.interactions.CreateModelInteraction;
import com.google.genai.gaos.models.interactions.GoogleSearch;
import com.google.genai.gaos.models.interactions.Interaction;
import com.google.genai.gaos.models.interactions.InteractionsInput;
import com.google.genai.gaos.models.interactions.Model;
import com.google.genai.gaos.models.interactions.ModelOutputStep;
import com.google.genai.gaos.models.interactions.Step;
import com.google.genai.gaos.models.interactions.TextContent;
import com.google.genai.gaos.models.interactions.URLCitation;
import com.google.genai.gaos.models.operations.CreateInteractionRequestBody;
import java.util.Arrays;
import java.util.Collections;

Client client = new Client();

CreateModelInteraction params =
    CreateModelInteraction.builder()
        .model(Model.of("gemini-3.8-flash"))
        .input(InteractionsInput.of("Who won the euro 2024?"))
        .tools(Arrays.asList(new GoogleSearch()))
        .build();

Interaction interaction =
    client.interactions.create(CreateInteractionRequestBody.of(params)).interaction().get();

System.out.println(interaction.outputText().orElse(""));

// Print citations
for (Step step : interaction.steps().orElse(Collections.emptyList())) {
  if (step instanceof ModelOutputStep outputStep) {
    for (Content contentBlock : outputStep.content().orElse(Collections.emptyList())) {
      if (contentBlock instanceof TextContent textContent && textContent.annotations().isPresent()) {
        System.out.println("\nCitations:");
        for (Annotation annotation : textContent.annotations().get()) {
          if (annotation instanceof URLCitation citation) {
            System.out.printf("  [%s](%s)%n", citation.title().orElse(""), citation.url().orElse(""));
          }
        }
      }
    }
  }
}

Go

package main

import (
    "context"
    "fmt"
    "log"

    "google.golang.org/genai"
    "google.golang.org/genai/interactions/models/interactions"
    "google.golang.org/genai/interactions/models/operations"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    res, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model: interactions.Model("gemini-3.8-flash"),
            Input: interactions.NewInteractionsInput("Calculate the 20th Fibonacci number and verify if it is prime."),
            Tools: []interactions.Tool{
                interactions.NewTool(interactions.CodeExecution{}),
            },
        }),
    })
    if err != nil {
        log.Fatal(err)
    }

    for _, step := range res.Interaction.Steps {
        if codeCall := step.CodeExecutionCallStep; codeCall != nil {
            fmt.Printf("Generated Code:\n%s\n", codeCall.Arguments.Code)
        } else if codeRes := step.CodeExecutionResultStep; codeRes != nil {
            fmt.Printf("Execution Output:\n%s\n", codeRes.Result)
        }
    }

    if res.Interaction.OutputText != nil {
        fmt.Println(*res.Interaction.OutputText)
    }
}

REST

curl -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemini-3.8-flash",
    "input": "Who won the euro 2024?",
    "tools": [{"type": "google_search"}]
  }'

Antwort

{
  "id": "v1_Chd...",
  "status": "completed",
  "steps": [
    {
      "type": "thought",
      "signature": "EvEFCu4F..."
    },
    {
      "type": "google_search_call",
      "arguments": {
        "queries": ["UEFA Euro 2024 winner"]
      }
    },
    {
      "type": "google_search_result",
      "call_id": "search_001",
      "result": [
        {
          "search_suggestions": "<!-- HTML and CSS search widget -->"
        }
      ]
    },
    {
      "type": "model_output",
      "content": [
        {
          "type": "text",
          "text": "Spain won Euro 2024, defeating England 2-1 in the final.",
          "annotations": [
            {
              "type": "url_citation",
              "url": "https://www.uefa.com/euro2024",
              "title": "uefa.com",
              "start_index": 0,
              "end_index": 56
            }
          ]
        }
      ]
    }
  ],
  "object": "interaction",
  "model": "gemini-3.8-flash",
}

Die Suchschritte werden im Interaktionsverlauf detailliert beschrieben und die endgültige Ausgabe enthält Inline-Zitationen, die auf Webquellen verweisen.

Informationen zum Extrahieren von Suchzitaten finden Sie im Leitfaden zur Fundierung mit der Google Suche. Informationen zum Kombinieren mehrerer Tools finden Sie im Leitfaden zur Kombination von Tools.

Code-Ausführung

Python-Code in einer sicheren Borg-Sandbox-Umgebung ausführen

URL-Kontext

Übergeben Sie öffentliche Web-URLs direkt, um Antworten auf Webseiteninhalte zu stützen.

Dateisuche

Hochgeladene Dokumente und Mediendateien indexieren und durchsuchen

Google Maps

Antworten mit realen Geodaten und Standortdaten fundieren.

Computernutzung

Browserautomatisierung und Bildschirminteraktion

9. Eigene Funktionen aufrufen

Mit Funktionsaufrufen können Sie das Modell mit Ihrem Code verbinden. Sie deklarieren den Namen und die Parameter einer Funktion, das Modell entscheidet, wann sie aufgerufen wird, und gibt strukturierte Argumente zurück. Sie führen sie lokal aus und senden das Ergebnis zurück.

Python

import json
from google import genai

client = genai.Client()

weather_tool = {
    "type": "function",
    "name": "get_current_temperature",
    "description": "Gets the current temperature for a given location.",
    "parameters": {
        "type": "object",
        "properties": {
            "location": {
                "type": "string",
                "description": "The city name, e.g. San Francisco",
            },
        },
        "required": ["location"],
    },
}

available_functions = {
    "get_current_temperature": lambda location: {
        "location": location, "temperature": "22", "unit": "celsius"
    },
}

user_input = "What is the temperature in London?"
previous_id = None

while True:
    interaction = client.interactions.create(
        model="gemini-3.8-flash",
        input=user_input,
        tools=[weather_tool],
        previous_interaction_id=previous_id,
    )

    function_results = []
    for step in interaction.steps:
        if step.type == "function_call":
            result = available_functions[step.name](**step.arguments)
            print(f"Called {step.name}({step.arguments}) → {result}")
            function_results.append({
                "type": "function_result",
                "name": step.name,
                "call_id": step.id,
                "result": [{"type": "text", "text": json.dumps(result)}],
            })

    if not function_results:
        break

    user_input = function_results
    previous_id = interaction.id

print(interaction.output_text)

JavaScript

import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({});

const weatherTool = {
  type: "function",
  name: "get_current_temperature",
  description: "Gets the current temperature for a given location.",
  parameters: {
    type: "object",
    properties: {
      location: {
        type: "string",
        description: "The city name, e.g. San Francisco",
      },
    },
    required: ["location"],
  },
};

const availableFunctions = {
  get_current_temperature: ({ location }) => ({
    location, temperature: "22", unit: "celsius"
  }),
};

let input = "What is the temperature in London?";
let previousId = null;
let interaction;

while (true) {
  interaction = await ai.interactions.create({
    model: "gemini-3.8-flash",
    input,
    tools: [weatherTool],
    previous_interaction_id: previousId,
  });

  const functionResults = [];
  for (const step of interaction.steps) {
    if (step.type === "function_call") {
      const result = availableFunctions[step.name](step.arguments);
      console.log(`Called ${step.name}(${JSON.stringify(step.arguments)}) →`, result);
      functionResults.push({
        type: "function_result",
        name: step.name,
        call_id: step.id,
        result: [{ type: "text", text: JSON.stringify(result) }],
      });
    }
  }

  if (functionResults.length === 0) break;

  input = functionResults;
  previousId = interaction.id;
}

console.log(interaction.output_text);

Java

import com.google.genai.Client;
import com.google.genai.gaos.models.interactions.CreateModelInteraction;
import com.google.genai.gaos.models.interactions.Function;
import com.google.genai.gaos.models.interactions.FunctionCallStep;
import com.google.genai.gaos.models.interactions.FunctionResultStep;
import com.google.genai.gaos.models.interactions.FunctionResultStepResultUnion;
import com.google.genai.gaos.models.interactions.Interaction;
import com.google.genai.gaos.models.interactions.InteractionsInput;
import com.google.genai.gaos.models.interactions.Model;
import com.google.genai.gaos.models.interactions.Step;
import com.google.genai.gaos.models.interactions.TextContent;
import com.google.genai.gaos.models.operations.CreateInteractionRequestBody;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collections;
import java.util.HashMap;
import java.util.List;
import java.util.Map;

Client client = new Client();

Map<String, Object> locationProp = new HashMap<>();
locationProp.put("type", "string");
locationProp.put("description", "The city name, e.g. San Francisco");

Map<String, Object> properties = new HashMap<>();
properties.put("location", locationProp);

Map<String, Object> parameters = new HashMap<>();
parameters.put("type", "object");
parameters.put("properties", properties);
parameters.put("required", Arrays.asList("location"));

Function weatherTool =
    Function.builder()
        .name("get_current_temperature")
        .description("Gets the current temperature for a given location.")
        .parameters(parameters)
        .build();

InteractionsInput userInput = InteractionsInput.of("What is the temperature in London?");
String previousId = null;
Interaction interaction = null;

while (true) {
  CreateModelInteraction.Builder paramsBuilder =
      CreateModelInteraction.builder()
          .model(Model.of("gemini-3.8-flash"))
          .input(userInput)
          .tools(Arrays.asList(weatherTool));
  if (previousId != null) {
    paramsBuilder.previousInteractionId(previousId);
  }

  interaction =
      client.interactions.create(CreateInteractionRequestBody.of(paramsBuilder.build())).interaction().get();

  List<Step> functionResults = new ArrayList<>();
  for (Step step : interaction.steps().orElse(Collections.emptyList())) {
    if (step instanceof FunctionCallStep fcStep) {
      String resultJson = "{\"location\": \"London\", \"temperature\": \"22\", \"unit\": \"celsius\"}";
      System.out.printf(
          "Called %s(%s) -> %s%n",
          fcStep.name().orElse(""), fcStep.arguments().orElse(Collections.emptyMap()), resultJson);
      functionResults.add(
          FunctionResultStep.builder()
              .name(fcStep.name().orElse(""))
              .callId(fcStep.id().orElse(""))
              .result(
                  FunctionResultStepResultUnion.of(
                      Arrays.asList(TextContent.builder().text(resultJson).build())))
              .build());
    }
  }

  if (functionResults.isEmpty()) {
    break;
  }

  userInput = InteractionsInput.ofStep(functionResults);
  previousId = interaction.id().orElse(null);
}

System.out.println(interaction.outputText().orElse(""));

Go

package main

import (
    "context"
    "fmt"
    "log"

    "google.golang.org/genai"
    "google.golang.org/genai/interactions/models/interactions"
    "google.golang.org/genai/interactions/models/operations"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    // Turn 1: Create a CSV file in the sandbox
    turn1, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model: interactions.Model("gemini-3.8-flash"),
            Input: interactions.NewInteractionsInput("Write a Python script to save a CSV file 'sales.csv' with 5 rows of sample data."),
            Tools: []interactions.Tool{
                interactions.NewTool(interactions.CodeExecution{}),
            },
        }),
    })
    if err != nil {
        log.Fatal(err)
    }

    var env *interactions.CreateModelInteractionEnvironment
    if turn1.Interaction.EnvironmentID != nil {
        env = genai.Ptr(interactions.NewCreateModelInteractionEnvironment(*turn1.Interaction.EnvironmentID))
    }

    // Turn 2: Reuse the sandbox environment to analyze the file created in Turn 1
    turn2, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model:                 interactions.Model("gemini-3.8-flash"),
            PreviousInteractionID: turn1.Interaction.ID,
            Environment:           env,
            Input:                 interactions.NewInteractionsInput("Now read 'sales.csv' and compute the total revenue."),
            Tools: []interactions.Tool{
                interactions.NewTool(interactions.CodeExecution{}),
            },
        }),
    })
    if err != nil {
        log.Fatal(err)
    }

    if turn2.Interaction.OutputText != nil {
        fmt.Println(*turn2.Interaction.OutputText)
    }
}

REST

# Turn 1: Send prompt with function declaration
RESPONSE1=$(curl -s -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemini-3.8-flash",
    "input": "What is the temperature in London?",
    "tools": [{
      "type": "function",
      "name": "get_current_temperature",
      "description": "Gets the current temperature for a given location.",
      "parameters": {
        "type": "object",
        "properties": {
          "location": {"type": "string", "description": "The city name"}
        },
        "required": ["location"]
      }
    }]
  }')

INTERACTION_ID=$(echo "$RESPONSE1" | jq -r '.id')
FC_NAME=$(echo "$RESPONSE1" | jq -r '.steps[] | select(.type=="function_call") | .name')
FC_ID=$(echo "$RESPONSE1" | jq -r '.steps[] | select(.type=="function_call") | .id')
echo "Function: $FC_NAME, Call ID: $FC_ID"

# Turn 2: Send function result back
curl -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemini-3.8-flash",
    "previous_interaction_id": "'$INTERACTION_ID'",
    "input": [{
      "type": "function_result",
      "name": "'$FC_NAME'",
      "call_id": "'$FC_ID'",
      "result": [{"type": "text", "text": "{\"location\": \"London\", \"temperature\": \"22\", \"unit\": \"celsius\"}"}]
    }],
    "tools": [{
      "type": "function",
      "name": "get_current_temperature",
      "description": "Gets the current temperature for a given location.",
      "parameters": {
        "type": "object",
        "properties": {
          "location": {"type": "string", "description": "The city name"}
        },
        "required": ["location"]
      }
    }]
  }'

Zustandslos

Sie können Funktionsaufrufe auch im statuslosen Modus verwenden, indem Sie den Unterhaltungsverlauf clientseitig verwalten und store=false festlegen. Im zustandslosen Modus müssen Sie den vollständigen Verlauf der Unterhaltung im Feld input jeder nachfolgenden Anfrage übergeben. Dieser Verlauf muss Folgendes enthalten:

  1. Der erste Schritt user_input.
  2. Alle vom Modell generierten Schritte, die in Turn 1 zurückgegeben werden (einschließlich der Schritte thought und function_call), werden genau so zurückgegeben, wie sie empfangen wurden.
  3. Der Schritt function_result, der die Ausgabe der ausgeführten Funktion enthält.

Python

import json
from google import genai

client = genai.Client()

weather_tool = {
    "type": "function",
    "name": "get_current_temperature",
    "description": "Gets the current temperature for a given location.",
    "parameters": {
        "type": "object",
        "properties": {
            "location": {
                "type": "string",
                "description": "The city name, e.g. San Francisco",
            },
        },
        "required": ["location"],
    },
}

available_functions = {
    "get_current_temperature": lambda location: {
        "location": location, "temperature": "22", "unit": "celsius"
    },
}

history = [
    {
        "type": "user_input",
        "content": [{"type": "text", "text": "What is the temperature in London?"}]
    }
]

while True:
    interaction = client.interactions.create(
        model="gemini-3.8-flash",
        store=False,
        input=history,
        tools=[weather_tool],
    )

    function_results = []
    for step in interaction.steps:
        history.append(step.model_dump())
        if step.type == "function_call":
            result = available_functions[step.name](**step.arguments)
            print(f"Called {step.name}({step.arguments}) → {result}")
            fn_result = {
                "type": "function_result",
                "name": step.name,
                "call_id": step.id,
                "result": [{"type": "text", "text": json.dumps(result)}],
            }
            function_results.append(fn_result)
            history.append(fn_result)

    if not function_results:
        break

print(interaction.output_text)

JavaScript

import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({});

const weatherTool = {
  type: "function",
  name: "get_current_temperature",
  description: "Gets the current temperature for a given location.",
  parameters: {
    type: "object",
    properties: {
      location: {
        type: "string",
        description: "The city name, e.g. San Francisco",
      },
    },
    required: ["location"],
  },
};

const availableFunctions = {
  get_current_temperature: ({ location }) => ({
    location, temperature: "22", unit: "celsius"
  }),
};

const history = [
  {
    type: "user_input",
    content: [{ type: "text", text: "What is the temperature in London?" }]
  }
];

let interaction;

while (true) {
  interaction = await ai.interactions.create({
    model: "gemini-3.8-flash",
    store: false,
    input: history,
    tools: [weatherTool],
  });

  const functionResults = [];
  for (const step of interaction.steps) {
    history.push(step);
    if (step.type === "function_call") {
      const result = availableFunctions[step.name](step.arguments);
      console.log(`Called ${step.name}(${JSON.stringify(step.arguments)}) →`, result);
      const fnResult = {
        type: "function_result",
        name: step.name,
        call_id: step.id,
        result: [{ type: "text", text: JSON.stringify(result) }],
      };
      functionResults.push(fnResult);
      history.push(fnResult);
    }
  }

  if (functionResults.length === 0) break;
}

console.log(interaction.output_text);

Java

import com.google.genai.Client;
import com.google.genai.gaos.models.interactions.CreateModelInteraction;
import com.google.genai.gaos.models.interactions.Function;
import com.google.genai.gaos.models.interactions.FunctionCallStep;
import com.google.genai.gaos.models.interactions.FunctionResultStep;
import com.google.genai.gaos.models.interactions.FunctionResultStepResultUnion;
import com.google.genai.gaos.models.interactions.Interaction;
import com.google.genai.gaos.models.interactions.InteractionsInput;
import com.google.genai.gaos.models.interactions.Model;
import com.google.genai.gaos.models.interactions.Step;
import com.google.genai.gaos.models.interactions.TextContent;
import com.google.genai.gaos.models.interactions.UserInputStep;
import com.google.genai.gaos.models.operations.CreateInteractionRequestBody;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collections;
import java.util.HashMap;
import java.util.List;
import java.util.Map;

Client client = new Client();

Map<String, Object> locationProp = new HashMap<>();
locationProp.put("type", "string");
locationProp.put("description", "The city name, e.g. San Francisco");

Map<String, Object> properties = new HashMap<>();
properties.put("location", locationProp);

Map<String, Object> parameters = new HashMap<>();
parameters.put("type", "object");
parameters.put("properties", properties);
parameters.put("required", Arrays.asList("location"));

Function weatherTool =
    Function.builder()
        .name("get_current_temperature")
        .description("Gets the current temperature for a given location.")
        .parameters(parameters)
        .build();

List<Step> history = new ArrayList<>();
history.add(
    UserInputStep.builder()
        .content(Arrays.asList(TextContent.builder().text("What is the temperature in London?").build()))
        .build());

Interaction interaction = null;

while (true) {
  CreateModelInteraction params =
      CreateModelInteraction.builder()
          .model(Model.of("gemini-3.8-flash"))
          .store(false)
          .input(InteractionsInput.ofStep(history))
          .tools(Arrays.asList(weatherTool))
          .build();

  interaction =
      client.interactions.create(CreateInteractionRequestBody.of(params)).interaction().get();

  List<Step> functionResults = new ArrayList<>();
  for (Step step : interaction.steps().orElse(Collections.emptyList())) {
    history.add(step);
    if (step instanceof FunctionCallStep fcStep) {
      String resultJson = "{\"location\": \"London\", \"temperature\": \"22\", \"unit\": \"celsius\"}";
      System.out.printf(
          "Called %s(%s) -> %s%n",
          fcStep.name().orElse(""), fcStep.arguments().orElse(Collections.emptyMap()), resultJson);
      FunctionResultStep fnResult =
          FunctionResultStep.builder()
              .name(fcStep.name().orElse(""))
              .callId(fcStep.id().orElse(""))
              .result(
                  FunctionResultStepResultUnion.of(
                      Arrays.asList(TextContent.builder().text(resultJson).build())))
              .build();
      functionResults.add(fnResult);
      history.add(fnResult);
    }
  }

  if (functionResults.isEmpty()) {
    break;
  }
}

System.out.println(interaction.outputText().orElse(""));

Go

package main

import (
    "context"
    "fmt"
    "log"

    "google.golang.org/genai"
    "google.golang.org/genai/interactions/models/interactions"
    "google.golang.org/genai/interactions/models/operations"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    recipeSchema := map[string]any{
        "type": "object",
        "properties": map[string]any{
            "recipe_name":       map[string]any{"type": "string"},
            "prep_time_minutes": map[string]any{"type": "integer"},
            "ingredients": map[string]any{
                "type":  "array",
                "items": map[string]any{"type": "string"},
            },
        },
        "required": []string{"recipe_name", "prep_time_minutes", "ingredients"},
    }

    res, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model: interactions.Model("gemini-3.8-flash"),
            Input: interactions.NewInteractionsInput("Give me a quick recipe for chocolate chip cookies."),
            ResponseFormat: genai.Ptr(interactions.NewCreateModelInteractionResponseFormat(
                interactions.NewResponseFormat(interactions.TextResponseFormat{
                    MimeType: interactions.TextResponseFormatMimeTypeApplicationJSON.ToPointer(),
                    Schema:   recipeSchema,
                }),
            )),
        }),
    })
    if err != nil {
        log.Fatal(err)
    }

    if res.Interaction.OutputText != nil {
        fmt.Println(*res.Interaction.OutputText)
    }
}

REST

# Turn 1: Send request with tools and store: false
RESPONSE1=$(curl -s -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemini-3.8-flash",
    "store": false,
    "input": [
      {
        "type": "user_input",
        "content": "What is the temperature in London?"
      }
    ],
    "tools": [{
      "type": "function",
      "name": "get_current_temperature",
      "description": "Gets the current temperature for a given location.",
      "parameters": {
        "type": "object",
        "properties": {
          "location": {"type": "string", "description": "The city name"}
        },
        "required": ["location"]
      }
    }]
  }')

# Extract model steps (thought, function_call)
MODEL_STEPS=$(echo "$RESPONSE1" | jq '.steps')
FC_NAME=$(echo "$RESPONSE1" | jq -r '.steps[] | select(.type=="function_call") | .name')
FC_ID=$(echo "$RESPONSE1" | jq -r '.steps[] | select(.type=="function_call") | .id')
echo "Function: $FC_NAME, Call ID: $FC_ID"

# Assume local execution returns:
RESULT="{\"location\": \"London\", \"temperature\": \"22\", \"unit\": \"celsius\"}"

# Reconstruct history for Turn 2
HISTORY=$(jq -n \
  --argjson first_input '[{"type": "user_input", "content": "What is the temperature in London?"}]' \
  --argjson model_steps "$MODEL_STEPS" \
  --arg fc_name "$FC_NAME" \
  --arg fc_id "$FC_ID" \
  --arg result "$RESULT" \
  '$first_input + $model_steps + [{"type": "function_result", "name": $fc_name, "call_id": $fc_id, "result": [{"type": "text", "text": $result}]}]')

# Turn 2: Send the full history
curl -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d "{
    \"model\": \"gemini-3.8-flash\",
    \"store\": false,
    \"input\": $HISTORY,
    \"tools\": [{
      \"type\": \"function\",
      \"name\": \"get_current_temperature\",
      \"description\": \"Gets the current temperature for a given location.\",
      \"parameters\": {
        \"type\": \"object\",
        \"properties\": {
          \"location\": {\"type\": \"string\", \"description\": \"The city name\"}
        },
        \"required\": [\"location\"]
      }
    }]
  }"

Antwort

In Runde 1 gibt das Modell eine Antwort mit dem Status requires_action und dem Schritt function_call zurück:

{
  "id": "v1_Chd...",
  "status": "requires_action",
  "steps": [
    {
      "type": "function_call",
      "id": "call_abc123",
      "name": "get_current_temperature",
      "arguments": {
        "location": "London"
      }
    }
  ],
  "object": "interaction",
  "model": "gemini-3.8-flash"
}

Nachdem Sie die Funktion lokal ausgeführt und das Ergebnis eingereicht haben (Turn 2), wird die endgültige abgeschlossene Interaktion zurückgegeben:

{
  "id": "v1_Chd...",
  "status": "completed",
  "steps": [
    {
      "type": "function_call",
      "id": "call_abc123",
      "name": "get_current_temperature",
      "arguments": {
        "location": "London"
      }
    },
    {
      "type": "model_output",
      "content": [
        {
          "type": "text",
          "text": "The temperature in London is currently 22°C."
        }
      ]
    }
  ],
  "object": "interaction",
  "model": "gemini-3.8-flash",
}

Weitere Informationen zu erweiterten Funktionen wie parallele Funktionsaufrufe oder Modi für die Funktionsauswahl finden Sie im Leitfaden zu Funktionsaufrufen.

10. Verwalteten Agent ausführen

Verwaltete Agents werden in einer Remote-Sandbox mit Zugriff auf Tools wie Codeausführung und Dateiverwaltung ausgeführt. Übergeben Sie ein agent anstelle eines model und legen Sie environment="remote" fest.

Python

from google import genai

client = genai.Client()

interaction = client.interactions.create(
    agent="antigravity-preview-09-2026",
    input="Write a Python script that generates the first 20 Fibonacci numbers and saves them to fibonacci.txt. Then read the file and print its contents.",
    environment="remote",
)
print(f"Environment: {interaction.environment_id}")
print(interaction.output_text)

JavaScript

import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({});

const interaction = await ai.interactions.create({
  agent: "antigravity-preview-09-2026",
  input: "Write a Python script that generates the first 20 Fibonacci numbers and saves them to fibonacci.txt. Then read the file and print its contents.",
  environment: "remote",
});
console.log(`Environment: ${interaction.environment_id}`);
console.log(interaction.output_text);

Java

import com.google.genai.Client;
import com.google.genai.gaos.models.interactions.CreateAgentInteraction;
import com.google.genai.gaos.models.interactions.CreateAgentInteractionEnvironment;
import com.google.genai.gaos.models.interactions.Interaction;
import com.google.genai.gaos.models.interactions.InteractionsInput;
import com.google.genai.gaos.models.operations.CreateInteractionRequestBody;

Client client = new Client();

CreateAgentInteraction params =
    CreateAgentInteraction.builder()
        .agent("antigravity-preview-09-2026")
        .input(
            InteractionsInput.of(
                "Write a Python script that generates the first 20 Fibonacci numbers and saves them to fibonacci.txt. Then read the file and print its contents."))
        .environment(CreateAgentInteractionEnvironment.of("remote"))
        .build();

Interaction interaction =
    client.interactions.create(CreateInteractionRequestBody.of(params)).interaction().get();
System.out.println("Environment: " + interaction.environmentId().orElse(""));
System.out.println(interaction.outputText().orElse(""));

Go

package main

import (
    "context"
    "fmt"
    "log"

    "google.golang.org/genai"
    "google.golang.org/genai/interactions/models/interactions"
    "google.golang.org/genai/interactions/models/operations"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    res, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateModelInteraction{
            Model: interactions.Model("gemini-3.8-pro"),
            Input: interactions.NewInteractionsInput("Solve this logic puzzle: Three gods A, B, and C are called True, False, and Random..."),
            GenerationConfig: &interactions.GenerationConfig{
                ThinkingLevel:     interactions.ThinkingLevelHigh.ToPointer(),
                ThinkingSummaries: interactions.ThinkingSummariesAuto.ToPointer(),
            },
        }),
    })
    if err != nil {
        log.Fatal(err)
    }

    // Print thought summaries if returned
    for _, step := range res.Interaction.Steps {
        if thought := step.ThoughtStep; thought != nil {
            for _, part := range thought.Summary {
                if part.TextContent != nil {
                    fmt.Printf("Thought Summary: %s\n", part.TextContent.Text)
                }
            }
        }
    }

    if res.Interaction.OutputText != nil {
        fmt.Printf("Answer: %s\n", *res.Interaction.OutputText)
    }
}

REST

curl -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "agent": "antigravity-preview-09-2026",
    "input": "Write a Python script that generates the first 20 Fibonacci numbers and saves them to fibonacci.txt. Then read the file and print its contents.",
    "environment": "remote"
  }'

Sie können auch benutzerdefinierte Agents mit eigenen Anweisungen, Skills und Datenquellen definieren und speichern.

Kurzanleitung

Ersten Agent-Aufruf starten, Antworten streamen und benutzerdefinierten Agenten erstellen

Antigravity-Agent

Funktionen, Tools, multimodale Eingabe und Preise für den Standard-Agenten.

KI-Agenten in AI Studio

Visuelle Umgebung zum Erstellen von Agent-Prototypen ohne Code.

11. Aufgaben im Hintergrund ausführen

Legen Sie background=True fest, um lange Aufgaben asynchron auszuführen. Rufen Sie die Ergebnisse mit interactions.get() ab. Weitere Informationen finden Sie im Leitfaden zur Ausführung im Hintergrund.

Python

import time
from google import genai

client = genai.Client()

interaction = client.interactions.create(
    model="gemini-3.8-flash",
    input="Write a detailed analysis of the impact of artificial intelligence on modern healthcare.",
    background=True,
)
print(f"Started background task: {interaction.id}")
print(f"Status: {interaction.status}")

# Poll for completion
while True:
    result = client.interactions.get(interaction.id)
    print(f"Status: {result.status}")
    if result.status == "completed":
        print(f"\nResult:\n{result.output_text}")
        break
    elif result.status == "failed":
        print(f"Failed: {result.error}")
        break
    time.sleep(5)

JavaScript

import { GoogleGenAI } from "@google/genai";

const ai = new GoogleGenAI({});

const interaction = await ai.interactions.create({
  model: "gemini-3.8-flash",
  input: "Write a detailed analysis of the impact of artificial intelligence on modern healthcare.",
  background: true,
});
console.log(`Started background task: ${interaction.id}`);
console.log(`Status: ${interaction.status}`);

// Poll for completion
while (true) {
  const result = await ai.interactions.get(interaction.id);
  console.log(`Status: ${result.status}`);
  if (result.status === "completed") {
    console.log(`\nResult:\n${result.output_text}`);
    break;
  } else if (result.status === "failed") {
    console.log(`Failed: ${result.error}`);
    break;
  }
  await new Promise(r => setTimeout(r, 5000));
}

Java

import com.google.genai.Client;
import com.google.genai.gaos.models.interactions.CreateModelInteraction;
import com.google.genai.gaos.models.interactions.Interaction;
import com.google.genai.gaos.models.interactions.InteractionStatus;
import com.google.genai.gaos.models.interactions.InteractionsInput;
import com.google.genai.gaos.models.interactions.Model;
import com.google.genai.gaos.models.operations.CreateInteractionRequestBody;
import com.google.genai.gaos.models.operations.GetInteractionByIdRequest;

Client client = new Client();

CreateModelInteraction params =
    CreateModelInteraction.builder()
        .model(Model.of("gemini-3.8-flash"))
        .input(
            InteractionsInput.of(
                "Write a detailed analysis of the impact of artificial intelligence on modern healthcare."))
        .background(true)
        .build();

Interaction interaction =
    client.interactions.create(CreateInteractionRequestBody.of(params)).interaction().get();
String interactionId = interaction.id().orElse("");
System.out.println("Started background task: " + interactionId);
System.out.println("Status: " + interaction.status().map(InteractionStatus::value).orElse(""));

// Poll for completion
while (true) {
  Interaction result =
      client.interactions.get(new GetInteractionByIdRequest(interactionId)).interaction().get();
  String status = result.status().map(InteractionStatus::value).orElse("");
  System.out.println("Status: " + status);
  if ("completed".equals(status)) {
    System.out.println("\nResult:\n" + result.outputText().orElse(""));
    break;
  } else if ("failed".equals(status)) {
    System.out.println("Failed: " + result.errors().orElse(null));
    break;
  }
  Thread.sleep(5000);
}

Go

package main

import (
    "context"
    "fmt"
    "log"
    "time"

    "google.golang.org/genai"
    "google.golang.org/genai/interactions/models/interactions"
    "google.golang.org/genai/interactions/models/operations"
)

func main() {
    ctx := context.Background()
    client, err := genai.NewClient(ctx, nil)
    if err != nil {
        log.Fatal(err)
    }

    // Start a Deep Research agent in the background
    res, err := client.Interactions.Create(ctx, operations.CreateInteractionRequest{
        Body: operations.NewCreateInteractionRequestBody(interactions.CreateAgentInteraction{
            Agent:      interactions.AgentOption("deep-research-pro-preview-12-2025"),
            Input:      interactions.NewInteractionsInput("Research the competitive landscape of solid-state EV batteries in 2026."),
            Background: genai.Ptr(true),
        }),
    })
    if err != nil {
        log.Fatal(err)
    }

    interaction := res.Interaction
    fmt.Printf("Started research job: %s\n", *interaction.ID)

    // Poll until completion
    for interaction.Status != interactions.InteractionStatusCompleted && interaction.Status != interactions.InteractionStatusFailed {
        time.Sleep(10 * time.Second)
        getRes, err := client.Interactions.Get(ctx, operations.GetInteractionByIDRequest{
            ID: *interaction.ID,
        })
        if err != nil {
            log.Fatal(err)
        }
        interaction = getRes.Interaction
        fmt.Printf("Current status: %s\n", interaction.Status)
    }

    if interaction.Status == interactions.InteractionStatusCompleted {
        if interaction.OutputText != nil {
            fmt.Println(*interaction.OutputText)
        }
    } else {
        fmt.Printf("Research failed: %v\n", interaction.Errors)
    }
}

REST

# Start a background task
RESPONSE=$(curl -s -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
  -H "x-goog-api-key: $GEMINI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemini-3.8-flash",
    "input": "Write a detailed analysis of the impact of artificial intelligence on modern healthcare.",
    "background": true
  }')

INTERACTION_ID=$(echo "$RESPONSE" | jq -r '.id')
echo "Started background task: $INTERACTION_ID"

# Poll for completion
while true; do
  RESULT=$(curl -s "https://generativelanguage.googleapis.com/v1beta/interactions/$INTERACTION_ID" \
    -H "x-goog-api-key: $GEMINI_API_KEY" \
    -H "Api-Revision: 2026-05-20")
  STATUS=$(echo "$RESULT" | jq -r '.status')
  echo "Status: $STATUS"
  if [ "$STATUS" = "completed" ]; then
    echo "$RESULT" | jq -r '.steps[] | select(.type=="model_output") | .content[] | select(.type=="text") | .text'
    break
  elif [ "$STATUS" = "failed" ]; then
    echo "Failed"
    break
  fi
  sleep 5
done

Antwort

Die erste Antwort wird sofort mit dem Status in_progress zurückgegeben:

{
  "id": "v1_abc123",
  "status": "in_progress",
  "object": "interaction",
  "model": "gemini-3.8-flash"
}

Nachdem die Hintergrundaufgabe vollständig ausgeführt wurde, wird beim Prüfen des Interaktionsstatus Folgendes zurückgegeben:

{
  "id": "v1_abc123",
  "status": "completed",
  "steps": [
    {
      "type": "model_output",
      "content": [
        {
          "type": "text",
          "text": "Artificial intelligence has transformed modern healthcare in several..."
        }
      ]
    }
  ],
  "object": "interaction",
  "model": "gemini-3.8-flash",
}

Informationen zum asynchronen Ausführen von Modellen und Agents finden Sie im Leitfaden zur Hintergrundausführung.

Nächste Schritte

  • Hintergrundausführung: Führen Sie lang andauernde Aufgaben asynchron aus und verwalten Sie den Status.
  • Textgenerierung: Systemanweisungen, Generierungskonfiguration und erweiterte Textmuster.
  • Bildgenerierung: Seitenverhältnisse, Bildbearbeitung und Stilreferenzen.
  • Bildverständnis: Klassifizierung, Objekterkennung und visuelle Fragen und Antworten.
  • Denken: Verwenden Sie die Chain-of-Thought-Methode für komplexe Aufgaben.
  • Funktionsaufrufe: Parallele, zusammengesetzte und eingeschränkte Funktionsmodi.
  • Google Suche: Fundierung, Zitate und Suchvorschläge.
  • Verwaltete KI-Agenten: Vordefinierte Agenten mit Code-Ausführung und Dateiverwaltung.
  • Deep Research: Autonome mehrstufige Recherche mit Planung und Synthese.
  • Strukturierte Ausgabe: JSON-Schemas, Enums und rekursive Typdefinitionen.