Eseguire la migrazione all'SDK Google GenAI

A partire dalla release di Gemini 2.0 alla fine del 2024, abbiamo introdotto un nuovo insieme di librerie chiamato SDK Google GenAI. Offre un'esperienza di sviluppo migliorata grazie a un'architettura client aggiornata e semplifica la transizione tra i flussi di lavoro per sviluppatori e aziende.

L'SDK Google GenAI è ora in disponibilità generale (GA) su tutte le piattaforme supportate. Se utilizzi una delle nostre librerie legacy, ti consigliamo vivamente di eseguire la migrazione.

Questa guida fornisce esempi di codice migrato prima e dopo per aiutarti a iniziare.

Installazione

Prima

Python

pip install -U -q "google-generativeai"

JavaScript

npm install @google/generative-ai

Vai

go get github.com/google/generative-ai-go

Dopo

Python

pip install -U -q "google-genai"

JavaScript

npm install @google/genai

Vai

go get google.golang.org/genai

Accesso API

Il vecchio SDK gestiva implicitamente il client API dietro le quinte utilizzando una serie di metodi ad hoc. In questo modo, era difficile gestire il client e le credenziali. Ora interagisci tramite un oggetto Client centrale. Questo oggetto Client funge da unico punto di ingresso per vari servizi API (ad es. models, chats, files, tunings), promuovendo la coerenza e semplificando la gestione delle credenziali e della configurazione in diverse chiamate API.

Prima (accesso all'API meno centralizzato)

Python

Il vecchio SDK non utilizzava esplicitamente un oggetto client di primo livello per la maggior parte delle chiamate API. Istanzieresti e interagiresti direttamente con gli oggetti GenerativeModel.

import google.generativeai as genai

# Directly create and use model objects
model = genai.GenerativeModel('gemini-1.5-flash')
response = model.generate_content(...)
chat = model.start_chat(...)

JavaScript

GoogleGenerativeAI era un punto centrale per i modelli e la chat, ma altre funzionalità come la gestione di file e cache spesso richiedevano l'importazione e l'istanziamento di classi client completamente separate.

import { GoogleGenerativeAI } from "@google/generative-ai";
import { GoogleAIFileManager, GoogleAICacheManager } from "@google/generative-ai/server"; // For files/caching

const genAI = new GoogleGenerativeAI("YOUR_API_KEY");
const fileManager = new GoogleAIFileManager("YOUR_API_KEY");
const cacheManager = new GoogleAICacheManager("YOUR_API_KEY");

// Get a model instance, then call methods on it
const model = genAI.getGenerativeModel({ model: "gemini-1.5-flash" });
const result = await model.generateContent(...);
const chat = model.startChat(...);

// Call methods on separate client objects for other services
const uploadedFile = await fileManager.uploadFile(...);
const cache = await cacheManager.create(...);

Vai

La funzione genai.NewClient ha creato un client, ma le operazioni del modello generativo venivano in genere chiamate su un'istanza GenerativeModel separata ottenuta da questo client. Altri servizi potrebbero essere stati accessibili tramite pacchetti o pattern distinti.

import (
      "github.com/google/generative-ai-go/genai"
      "github.com/google/generative-ai-go/genai/fileman" // For files
      "google.golang.org/api/option"
)

client, err := genai.NewClient(ctx, option.WithAPIKey("YOUR_API_KEY"))
fileClient, err := fileman.NewClient(ctx, option.WithAPIKey("YOUR_API_KEY"))

// Get a model instance, then call methods on it
model := client.GenerativeModel("gemini-1.5-flash")
resp, err := model.GenerateContent(...)
cs := model.StartChat()

// Call methods on separate client objects for other services
uploadedFile, err := fileClient.UploadFile(...)

Dopo (oggetto cliente centralizzato)

Python

from google import genai

# Create a single client object
client = genai.Client()

# Access API methods through services on the client object
response = client.models.generate_content(...)
chat = client.chats.create(...)
my_file = client.files.upload(...)
tuning_job = client.tunings.tune(...)

JavaScript

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

// Create a single client object
const ai = new GoogleGenAI({apiKey: "YOUR_API_KEY"});

// Access API methods through services on the client object
const response = await ai.models.generateContent(...);
const chat = ai.chats.create(...);
const uploadedFile = await ai.files.upload(...);
const cache = await ai.caches.create(...);

Vai

import "google.golang.org/genai"

// Create a single client object
client, err := genai.NewClient(ctx, nil)

// Access API methods through services on the client object
result, err := client.Models.GenerateContent(...)
chat, err := client.Chats.Create(...)
uploadedFile, err := client.Files.Upload(...)
tuningJob, err := client.Tunings.Tune(...)

Autenticazione

Sia le librerie precedenti sia quelle nuove eseguono l'autenticazione utilizzando le chiavi API. Puoi creare la tua chiave API in Google AI Studio.

Prima

Python

Il vecchio SDK gestiva implicitamente l'oggetto client API.

import google.generativeai as genai

genai.configure(api_key=...)

JavaScript

import { GoogleGenerativeAI } from "@google/generative-ai";

const genAI = new GoogleGenerativeAI("GOOGLE_API_KEY");

Vai

Importa le librerie Google:

import (
      "github.com/google/generative-ai-go/genai"
      "google.golang.org/api/option"
)

Crea il client:

client, err := genai.NewClient(ctx, option.WithAPIKey("GOOGLE_API_KEY"))

Dopo

Python

Con Google GenAI SDK, devi prima creare un client API, che viene utilizzato per chiamare l'API. Il nuovo SDK recupererà la chiave API da una delle variabili di ambiente GEMINI_API_KEY o GOOGLE_API_KEY se non ne passi una al client.

export GEMINI_API_KEY="YOUR_API_KEY"
from google import genai

client = genai.Client() # Set the API key using the GEMINI_API_KEY env var.
                        # Alternatively, you could set the API key explicitly:
                        # client = genai.Client(api_key="your_api_key")

JavaScript

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

const ai = new GoogleGenAI({apiKey: "GEMINI_API_KEY"});

Vai

Importa la libreria GenAI:

import "google.golang.org/genai"

Crea il client:

client, err := genai.NewClient(ctx, &genai.ClientConfig{
        Backend:  genai.BackendGeminiAPI,
})

Generazione di contenuti

Testo

Prima

Python

In precedenza, non esistevano oggetti client e si accedeva alle API direttamente tramite oggetti GenerativeModel.

import google.generativeai as genai

model = genai.GenerativeModel('gemini-1.5-flash')
response = model.generate_content(
    'Tell me a story in 300 words'
)
print(response.text)

JavaScript

import { GoogleGenerativeAI } from "@google/generative-ai";

const genAI = new GoogleGenerativeAI(process.env.API_KEY);
const model = genAI.getGenerativeModel({ model: "gemini-1.5-flash" });
const prompt = "Tell me a story in 300 words";

const result = await model.generateContent(prompt);
console.log(result.response.text());

Vai

ctx := context.Background()
client, err := genai.NewClient(ctx, option.WithAPIKey("GOOGLE_API_KEY"))
if err != nil {
    log.Fatal(err)
}
defer client.Close()

model := client.GenerativeModel("gemini-1.5-flash")
resp, err := model.GenerateContent(ctx, genai.Text("Tell me a story in 300 words."))
if err != nil {
    log.Fatal(err)
}

printResponse(resp) // utility for printing response parts

Dopo

Python

Il nuovo SDK Google GenAI fornisce l'accesso a tutti i metodi API tramite l'oggetto Client. Ad eccezione di alcuni casi speciali con stato (chat e live-api session), si tratta di funzioni senza stato. Per utilità e uniformità, gli oggetti restituiti sono classi pydantic.

from google import genai
client = genai.Client()

response = client.models.generate_content(
    model='gemini-2.0-flash',
    contents='Tell me a story in 300 words.'
)
print(response.text)

print(response.model_dump_json(
    exclude_none=True, indent=4))

JavaScript

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

const ai = new GoogleGenAI({ apiKey: "GOOGLE_API_KEY" });

const response = await ai.models.generateContent({
  model: "gemini-2.0-flash",
  contents: "Tell me a story in 300 words.",
});
console.log(response.text);

Vai

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

result, err := client.Models.GenerateContent(ctx, "gemini-2.0-flash", genai.Text("Tell me a story in 300 words."), nil)
if err != nil {
    log.Fatal(err)
}
debugPrint(result) // utility for printing result

Immagine

Prima

Python

import google.generativeai as genai

model = genai.GenerativeModel('gemini-1.5-flash')
response = model.generate_content([
    'Tell me a story based on this image',
    Image.open(image_path)
])
print(response.text)

JavaScript

import { GoogleGenerativeAI } from "@google/generative-ai";

const genAI = new GoogleGenerativeAI("GOOGLE_API_KEY");
const model = genAI.getGenerativeModel({ model: "gemini-1.5-flash" });

function fileToGenerativePart(path, mimeType) {
  return {
    inlineData: {
      data: Buffer.from(fs.readFileSync(path)).toString("base64"),
      mimeType,
    },
  };
}

const prompt = "Tell me a story based on this image";

const imagePart = fileToGenerativePart(
  `path/to/organ.jpg`,
  "image/jpeg",
);

const result = await model.generateContent([prompt, imagePart]);
console.log(result.response.text());

Vai

ctx := context.Background()
client, err := genai.NewClient(ctx, option.WithAPIKey("GOOGLE_API_KEY"))
if err != nil {
    log.Fatal(err)
}
defer client.Close()

model := client.GenerativeModel("gemini-1.5-flash")

imgData, err := os.ReadFile("path/to/organ.jpg")
if err != nil {
    log.Fatal(err)
}

resp, err := model.GenerateContent(ctx,
    genai.Text("Tell me about this instrument"),
    genai.ImageData("jpeg", imgData))
if err != nil {
    log.Fatal(err)
}

printResponse(resp) // utility for printing response

Dopo

Python

Molte delle stesse funzionalità di convenienza sono presenti nel nuovo SDK. Ad esempio, PIL.Image oggetti vengono convertiti automaticamente.

from google import genai
from PIL import Image

client = genai.Client()

response = client.models.generate_content(
    model='gemini-2.0-flash',
    contents=[
        'Tell me a story based on this image',
        Image.open(image_path)
    ]
)
print(response.text)

JavaScript

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

const ai = new GoogleGenAI({ apiKey: "GOOGLE_API_KEY" });

const organ = await ai.files.upload({
  file: "path/to/organ.jpg",
});

const response = await ai.models.generateContent({
  model: "gemini-2.0-flash",
  contents: [
    createUserContent([
      "Tell me a story based on this image",
      createPartFromUri(organ.uri, organ.mimeType)
    ]),
  ],
});
console.log(response.text);

Vai

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

imgData, err := os.ReadFile("path/to/organ.jpg")
if err != nil {
    log.Fatal(err)
}

parts := []*genai.Part{
    {Text: "Tell me a story based on this image"},
    {InlineData: &genai.Blob{Data: imgData, MIMEType: "image/jpeg"}},
}
contents := []*genai.Content{
    {Parts: parts},
}

result, err := client.Models.GenerateContent(ctx, "gemini-2.0-flash", contents, nil)
if err != nil {
    log.Fatal(err)
}
debugPrint(result) // utility for printing result

Streaming

Prima

Python

import google.generativeai as genai

response = model.generate_content(
    "Write a cute story about cats.",
    stream=True)
for chunk in response:
    print(chunk.text)

JavaScript

import { GoogleGenerativeAI } from "@google/generative-ai";

const genAI = new GoogleGenerativeAI("GOOGLE_API_KEY");
const model = genAI.getGenerativeModel({ model: "gemini-1.5-flash" });

const prompt = "Write a story about a magic backpack.";

const result = await model.generateContentStream(prompt);

// Print text as it comes in.
for await (const chunk of result.stream) {
  const chunkText = chunk.text();
  process.stdout.write(chunkText);
}

Vai

ctx := context.Background()
client, err := genai.NewClient(ctx, option.WithAPIKey("GOOGLE_API_KEY"))
if err != nil {
    log.Fatal(err)
}
defer client.Close()

model := client.GenerativeModel("gemini-1.5-flash")
iter := model.GenerateContentStream(ctx, genai.Text("Write a story about a magic backpack."))
for {
    resp, err := iter.Next()
    if err == iterator.Done {
        break
    }
    if err != nil {
        log.Fatal(err)
    }
    printResponse(resp) // utility for printing the response
}

Dopo

Python

from google import genai

client = genai.Client()

for chunk in client.models.generate_content_stream(
  model='gemini-2.0-flash',
  contents='Tell me a story in 300 words.'
):
    print(chunk.text)

JavaScript

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

const ai = new GoogleGenAI({ apiKey: "GOOGLE_API_KEY" });

const response = await ai.models.generateContentStream({
  model: "gemini-2.0-flash",
  contents: "Write a story about a magic backpack.",
});
let text = "";
for await (const chunk of response) {
  console.log(chunk.text);
  text += chunk.text;
}

Vai

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

for result, err := range client.Models.GenerateContentStream(
    ctx,
    "gemini-2.0-flash",
    genai.Text("Write a story about a magic backpack."),
    nil,
) {
    if err != nil {
        log.Fatal(err)
    }
    fmt.Print(result.Candidates[0].Content.Parts[0].Text)
}

Configurazione

Prima

Python

import google.generativeai as genai

model = genai.GenerativeModel(
  'gemini-1.5-flash',
    system_instruction='you are a story teller for kids under 5 years old',
    generation_config=genai.GenerationConfig(
      max_output_tokens=400,
      top_k=2,
      top_p=0.5,
      temperature=0.5,
      response_mime_type='application/json',
      stop_sequences=['\n'],
    )
)
response = model.generate_content('tell me a story in 100 words')

JavaScript

import { GoogleGenerativeAI } from "@google/generative-ai";

const genAI = new GoogleGenerativeAI("GOOGLE_API_KEY");
const model = genAI.getGenerativeModel({
  model: "gemini-1.5-flash",
  generationConfig: {
    candidateCount: 1,
    stopSequences: ["x"],
    maxOutputTokens: 20,
    temperature: 1.0,
  },
});

const result = await model.generateContent(
  "Tell me a story about a magic backpack.",
);
console.log(result.response.text())

Vai

ctx := context.Background()
client, err := genai.NewClient(ctx, option.WithAPIKey("GOOGLE_API_KEY"))
if err != nil {
    log.Fatal(err)
}
defer client.Close()

model := client.GenerativeModel("gemini-1.5-flash")
model.SetTemperature(0.5)
model.SetTopP(0.5)
model.SetTopK(2.0)
model.SetMaxOutputTokens(100)
model.ResponseMIMEType = "application/json"
resp, err := model.GenerateContent(ctx, genai.Text("Tell me about New York"))
if err != nil {
    log.Fatal(err)
}
printResponse(resp) // utility for printing response

Dopo

Python

Per tutti i metodi nel nuovo SDK, gli argomenti obbligatori vengono forniti come argomenti di parole chiave. Tutti gli input facoltativi vengono forniti nell'argomento config. Gli argomenti di configurazione possono essere specificati come dizionari Python o classi Config nello spazio dei nomi google.genai.types. Per utilità e uniformità, tutte le definizioni all'interno del modulo types sono classi pydantic.

from google import genai
from google.genai import types

client = genai.Client()

response = client.models.generate_content(
  model='gemini-2.0-flash',
  contents='Tell me a story in 100 words.',
  config=types.GenerateContentConfig(
      system_instruction='you are a story teller for kids under 5 years old',
      max_output_tokens= 400,
      top_k= 2,
      top_p= 0.5,
      temperature= 0.5,
      response_mime_type= 'application/json',
      stop_sequences= ['\n'],
      seed=42,
  ),
)

JavaScript

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

const ai = new GoogleGenAI({ apiKey: "GOOGLE_API_KEY" });

const response = await ai.models.generateContent({
  model: "gemini-2.0-flash",
  contents: "Tell me a story about a magic backpack.",
  config: {
    candidateCount: 1,
    stopSequences: ["x"],
    maxOutputTokens: 20,
    temperature: 1.0,
  },
});

console.log(response.text);

Vai

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

result, err := client.Models.GenerateContent(ctx,
    "gemini-2.0-flash",
    genai.Text("Tell me about New York"),
    &genai.GenerateContentConfig{
        Temperature:      genai.Ptr[float32](0.5),
        TopP:             genai.Ptr[float32](0.5),
        TopK:             genai.Ptr[float32](2.0),
        ResponseMIMEType: "application/json",
        StopSequences:    []string{"Yankees"},
        CandidateCount:   2,
        Seed:             genai.Ptr[int32](42),
        MaxOutputTokens:  128,
        PresencePenalty:  genai.Ptr[float32](0.5),
        FrequencyPenalty: genai.Ptr[float32](0.5),
    },
)
if err != nil {
    log.Fatal(err)
}
debugPrint(result) // utility for printing response

Impostazioni di sicurezza

Genera una risposta con le impostazioni di sicurezza:

Prima

Python

import google.generativeai as genai

model = genai.GenerativeModel('gemini-1.5-flash')
response = model.generate_content(
    'say something bad',
    safety_settings={
        'HATE': 'BLOCK_ONLY_HIGH',
        'HARASSMENT': 'BLOCK_ONLY_HIGH',
  }
)

JavaScript

import { GoogleGenerativeAI, HarmCategory, HarmBlockThreshold } from "@google/generative-ai";

const genAI = new GoogleGenerativeAI("GOOGLE_API_KEY");
const model = genAI.getGenerativeModel({
  model: "gemini-1.5-flash",
  safetySettings: [
    {
      category: HarmCategory.HARM_CATEGORY_HARASSMENT,
      threshold: HarmBlockThreshold.BLOCK_LOW_AND_ABOVE,
    },
  ],
});

const unsafePrompt =
  "I support Martians Soccer Club and I think " +
  "Jupiterians Football Club sucks! Write an ironic phrase telling " +
  "them how I feel about them.";

const result = await model.generateContent(unsafePrompt);

try {
  result.response.text();
} catch (e) {
  console.error(e);
  console.log(result.response.candidates[0].safetyRatings);
}

Dopo

Python

from google import genai
from google.genai import types

client = genai.Client()

response = client.models.generate_content(
  model='gemini-2.0-flash',
  contents='say something bad',
  config=types.GenerateContentConfig(
      safety_settings= [
          types.SafetySetting(
              category='HARM_CATEGORY_HATE_SPEECH',
              threshold='BLOCK_ONLY_HIGH'
          ),
      ]
  ),
)

JavaScript

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

const ai = new GoogleGenAI({ apiKey: "GOOGLE_API_KEY" });
const unsafePrompt =
  "I support Martians Soccer Club and I think " +
  "Jupiterians Football Club sucks! Write an ironic phrase telling " +
  "them how I feel about them.";

const response = await ai.models.generateContent({
  model: "gemini-2.0-flash",
  contents: unsafePrompt,
  config: {
    safetySettings: [
      {
        category: "HARM_CATEGORY_HARASSMENT",
        threshold: "BLOCK_ONLY_HIGH",
      },
    ],
  },
});

console.log("Finish reason:", response.candidates[0].finishReason);
console.log("Safety ratings:", response.candidates[0].safetyRatings);

Asinc

Prima

Python

import google.generativeai as genai

model = genai.GenerativeModel('gemini-1.5-flash')
response = model.generate_content_async(
    'tell me a story in 100 words'
)

Dopo

Python

Per utilizzare il nuovo SDK con asyncio, esiste un'implementazione separata di ogni metodo in async in client.aio.

from google import genai

client = genai.Client()

response = await client.aio.models.generate_content(
    model='gemini-2.0-flash',
    contents='Tell me a story in 300 words.'
)

Chat

Avvia una chat e invia un messaggio al modello:

Prima

Python

import google.generativeai as genai

model = genai.GenerativeModel('gemini-1.5-flash')
chat = model.start_chat()

response = chat.send_message(
    "Tell me a story in 100 words")
response = chat.send_message(
    "What happened after that?")

JavaScript

import { GoogleGenerativeAI } from "@google/generative-ai";

const genAI = new GoogleGenerativeAI("GOOGLE_API_KEY");
const model = genAI.getGenerativeModel({ model: "gemini-1.5-flash" });
const chat = model.startChat({
  history: [
    {
      role: "user",
      parts: [{ text: "Hello" }],
    },
    {
      role: "model",
      parts: [{ text: "Great to meet you. What would you like to know?" }],
    },
  ],
});
let result = await chat.sendMessage("I have 2 dogs in my house.");
console.log(result.response.text());
result = await chat.sendMessage("How many paws are in my house?");
console.log(result.response.text());

Vai

ctx := context.Background()
client, err := genai.NewClient(ctx, option.WithAPIKey("GOOGLE_API_KEY"))
if err != nil {
    log.Fatal(err)
}
defer client.Close()

model := client.GenerativeModel("gemini-1.5-flash")
cs := model.StartChat()

cs.History = []*genai.Content{
    {
        Parts: []genai.Part{
            genai.Text("Hello, I have 2 dogs in my house."),
        },
        Role: "user",
    },
    {
        Parts: []genai.Part{
            genai.Text("Great to meet you. What would you like to know?"),
        },
        Role: "model",
    },
}

res, err := cs.SendMessage(ctx, genai.Text("How many paws are in my house?"))
if err != nil {
    log.Fatal(err)
}
printResponse(res) // utility for printing the response

Dopo

Python

from google import genai

client = genai.Client()

chat = client.chats.create(model='gemini-2.0-flash')

response = chat.send_message(
    message='Tell me a story in 100 words')
response = chat.send_message(
    message='What happened after that?')

JavaScript

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

const ai = new GoogleGenAI({ apiKey: "GOOGLE_API_KEY" });
const chat = ai.chats.create({
  model: "gemini-2.0-flash",
  history: [
    {
      role: "user",
      parts: [{ text: "Hello" }],
    },
    {
      role: "model",
      parts: [{ text: "Great to meet you. What would you like to know?" }],
    },
  ],
});

const response1 = await chat.sendMessage({
  message: "I have 2 dogs in my house.",
});
console.log("Chat response 1:", response1.text);

const response2 = await chat.sendMessage({
  message: "How many paws are in my house?",
});
console.log("Chat response 2:", response2.text);

Vai

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

chat, err := client.Chats.Create(ctx, "gemini-2.0-flash", nil, nil)
if err != nil {
    log.Fatal(err)
}

result, err := chat.SendMessage(ctx, genai.Part{Text: "Hello, I have 2 dogs in my house."})
if err != nil {
    log.Fatal(err)
}
debugPrint(result) // utility for printing result

result, err = chat.SendMessage(ctx, genai.Part{Text: "How many paws are in my house?"})
if err != nil {
    log.Fatal(err)
}
debugPrint(result) // utility for printing result

Chiamata di funzione

Prima

Python

import google.generativeai as genai
from enum import Enum

def get_current_weather(location: str) -> str:
    """Get the current whether in a given location.

    Args:
        location: required, The city and state, e.g. San Franciso, CA
        unit: celsius or fahrenheit
    """
    print(f'Called with: {location=}')
    return "23C"

model = genai.GenerativeModel(
    model_name="gemini-1.5-flash",
    tools=[get_current_weather]
)

response = model.generate_content("What is the weather in San Francisco?")
function_call = response.candidates[0].parts[0].function_call

Dopo

Python

Nel nuovo SDK, la chiamata automatica delle funzioni è l'impostazione predefinita. Qui puoi disattivarla.

from google import genai
from google.genai import types

client = genai.Client()

def get_current_weather(location: str) -> str:
    """Get the current whether in a given location.

    Args:
        location: required, The city and state, e.g. San Franciso, CA
        unit: celsius or fahrenheit
    """
    print(f'Called with: {location=}')
    return "23C"

response = client.models.generate_content(
  model='gemini-2.0-flash',
  contents="What is the weather like in Boston?",
  config=types.GenerateContentConfig(
      tools=[get_current_weather],
      automatic_function_calling={'disable': True},
  ),
)

function_call = response.candidates[0].content.parts[0].function_call

Chiamata automatica di funzioni

Prima

Python

La vecchia SDK supporta solo la chiamata automatica di funzioni nella chat. Nel nuovo SDK questo è il comportamento predefinito in generate_content.

import google.generativeai as genai

def get_current_weather(city: str) -> str:
    return "23C"

model = genai.GenerativeModel(
    model_name="gemini-1.5-flash",
    tools=[get_current_weather]
)

chat = model.start_chat(
    enable_automatic_function_calling=True)
result = chat.send_message("What is the weather in San Francisco?")

Dopo

Python

from google import genai
from google.genai import types
client = genai.Client()

def get_current_weather(city: str) -> str:
    return "23C"

response = client.models.generate_content(
  model='gemini-2.0-flash',
  contents="What is the weather like in Boston?",
  config=types.GenerateContentConfig(
      tools=[get_current_weather]
  ),
)

Esecuzione del codice

L'esecuzione del codice è uno strumento che consente al modello di generare codice Python, eseguirlo e restituire il risultato.

Prima

Python

import google.generativeai as genai

model = genai.GenerativeModel(
    model_name="gemini-1.5-flash",
    tools="code_execution"
)

result = model.generate_content(
  "What is the sum of the first 50 prime numbers? Generate and run code for "
  "the calculation, and make sure you get all 50.")

JavaScript

import { GoogleGenerativeAI } from "@google/generative-ai";

const genAI = new GoogleGenerativeAI("GOOGLE_API_KEY");
const model = genAI.getGenerativeModel({
  model: "gemini-1.5-flash",
  tools: [{ codeExecution: {} }],
});

const result = await model.generateContent(
  "What is the sum of the first 50 prime numbers? " +
    "Generate and run code for the calculation, and make sure you get " +
    "all 50.",
);

console.log(result.response.text());

Dopo

Python

from google import genai
from google.genai import types

client = genai.Client()

response = client.models.generate_content(
    model='gemini-2.0-flash',
    contents='What is the sum of the first 50 prime numbers? Generate and run '
            'code for the calculation, and make sure you get all 50.',
    config=types.GenerateContentConfig(
        tools=[types.Tool(code_execution=types.ToolCodeExecution)],
    ),
)

JavaScript

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

const ai = new GoogleGenAI({ apiKey: "GOOGLE_API_KEY" });

const response = await ai.models.generateContent({
  model: "gemini-2.0-pro-exp-02-05",
  contents: `Write and execute code that calculates the sum of the first 50 prime numbers.
            Ensure that only the executable code and its resulting output are generated.`,
});

// Each part may contain text, executable code, or an execution result.
for (const part of response.candidates[0].content.parts) {
  console.log(part);
  console.log("\n");
}

console.log("-".repeat(80));
// The `.text` accessor concatenates the parts into a markdown-formatted text.
console.log("\n", response.text);

Fondatezza della Ricerca

GoogleSearch (Gemini>=2.0) e GoogleSearchRetrieval (Gemini < 2.0) sono strumenti che consentono al modello di recuperare dati web pubblici per il grounding, basati su Google.

Prima

Python

import google.generativeai as genai

model = genai.GenerativeModel('gemini-1.5-flash')
response = model.generate_content(
    contents="what is the Google stock price?",
    tools='google_search_retrieval'
)

Dopo

Python

from google import genai
from google.genai import types

client = genai.Client()

response = client.models.generate_content(
    model='gemini-2.0-flash',
    contents='What is the Google stock price?',
    config=types.GenerateContentConfig(
        tools=[
            types.Tool(
                google_search=types.GoogleSearch()
            )
        ]
    )
)

Risposta JSON

Genera risposte in formato JSON.

Prima

Python

Specificando un response_schema e impostando response_mime_type="application/json", gli utenti possono vincolare il modello a produrre una risposta JSON seguendo una determinata struttura.

import google.generativeai as genai
import typing_extensions as typing

class CountryInfo(typing.TypedDict):
    name: str
    population: int
    capital: str
    continent: str
    major_cities: list[str]
    gdp: int
    official_language: str
    total_area_sq_mi: int

model = genai.GenerativeModel(model_name="gemini-1.5-flash")
result = model.generate_content(
    "Give me information of the United States",
    generation_config=genai.GenerationConfig(
        response_mime_type="application/json",
        response_schema = CountryInfo
    ),
)

JavaScript

import { GoogleGenerativeAI, SchemaType } from "@google/generative-ai";

const genAI = new GoogleGenerativeAI("GOOGLE_API_KEY");

const schema = {
  description: "List of recipes",
  type: SchemaType.ARRAY,
  items: {
    type: SchemaType.OBJECT,
    properties: {
      recipeName: {
        type: SchemaType.STRING,
        description: "Name of the recipe",
        nullable: false,
      },
    },
    required: ["recipeName"],
  },
};

const model = genAI.getGenerativeModel({
  model: "gemini-1.5-pro",
  generationConfig: {
    responseMimeType: "application/json",
    responseSchema: schema,
  },
});

const result = await model.generateContent(
  "List a few popular cookie recipes.",
);
console.log(result.response.text());

Dopo

Python

Il nuovo SDK utilizza le classi pydantic per fornire lo schema (anche se puoi passare un genai.types.Schema o un dict equivalente). Quando possibile, l'SDK analizzerà il JSON restituito e restituirà il risultato in response.parsed. Se hai fornito una classe pydantic come schema, l'SDK convertirà JSON in un'istanza della classe.

from google import genai
from pydantic import BaseModel

client = genai.Client()

class CountryInfo(BaseModel):
    name: str
    population: int
    capital: str
    continent: str
    major_cities: list[str]
    gdp: int
    official_language: str
    total_area_sq_mi: int

response = client.models.generate_content(
    model='gemini-2.0-flash',
    contents='Give me information of the United States.',
    config={
        'response_mime_type': 'application/json',
        'response_schema': CountryInfo,
    },
)

response.parsed

JavaScript

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

const ai = new GoogleGenAI({ apiKey: "GOOGLE_API_KEY" });
const response = await ai.models.generateContent({
  model: "gemini-2.0-flash",
  contents: "List a few popular cookie recipes.",
  config: {
    responseMimeType: "application/json",
    responseSchema: {
      type: "array",
      items: {
        type: "object",
        properties: {
          recipeName: { type: "string" },
          ingredients: { type: "array", items: { type: "string" } },
        },
        required: ["recipeName", "ingredients"],
      },
    },
  },
});
console.log(response.text);

File

Carica

Caricare un file:

Prima

Python

import requests
import pathlib
import google.generativeai as genai

# Download file
response = requests.get(
    'https://storage.googleapis.com/generativeai-downloads/data/a11.txt')
pathlib.Path('a11.txt').write_text(response.text)

file = genai.upload_file(path='a11.txt')

model = genai.GenerativeModel('gemini-1.5-flash')
response = model.generate_content([
    'Can you summarize this file:',
    my_file
])
print(response.text)

Dopo

Python

import requests
import pathlib
from google import genai

client = genai.Client()

# Download file
response = requests.get(
    'https://storage.googleapis.com/generativeai-downloads/data/a11.txt')
pathlib.Path('a11.txt').write_text(response.text)

my_file = client.files.upload(file='a11.txt')

response = client.models.generate_content(
    model='gemini-2.0-flash',
    contents=[
        'Can you summarize this file:',
        my_file
    ]
)
print(response.text)

Elenca e ottieni

Elenca i file caricati e ottieni un file caricato con un nome file:

Prima

Python

import google.generativeai as genai

for file in genai.list_files():
  print(file.name)

file = genai.get_file(name=file.name)

Dopo

Python

from google import genai
client = genai.Client()

for file in client.files.list():
    print(file.name)

file = client.files.get(name=file.name)

Elimina

Eliminare un file:

Prima

Python

import pathlib
import google.generativeai as genai

pathlib.Path('dummy.txt').write_text(dummy)
dummy_file = genai.upload_file(path='dummy.txt')

file = genai.delete_file(name=dummy_file.name)

Dopo

Python

import pathlib
from google import genai

client = genai.Client()

pathlib.Path('dummy.txt').write_text(dummy)
dummy_file = client.files.upload(file='dummy.txt')

response = client.files.delete(name=dummy_file.name)

Memorizzazione nella cache del contesto

La memorizzazione nella cache del contesto consente all'utente di passare i contenuti al modello una sola volta, memorizzare nella cache i token di input e poi fare riferimento ai token memorizzati nella cache nelle chiamate successive per ridurre i costi.

Prima

Python

import requests
import pathlib
import google.generativeai as genai
from google.generativeai import caching

# Download file
response = requests.get(
    'https://storage.googleapis.com/generativeai-downloads/data/a11.txt')
pathlib.Path('a11.txt').write_text(response.text)

# Upload file
document = genai.upload_file(path="a11.txt")

# Create cache
apollo_cache = caching.CachedContent.create(
    model="gemini-1.5-flash-001",
    system_instruction="You are an expert at analyzing transcripts.",
    contents=[document],
)

# Generate response
apollo_model = genai.GenerativeModel.from_cached_content(
    cached_content=apollo_cache
)
response = apollo_model.generate_content("Find a lighthearted moment from this transcript")

JavaScript

import { GoogleAICacheManager, GoogleAIFileManager } from "@google/generative-ai/server";
import { GoogleGenerativeAI } from "@google/generative-ai";

const cacheManager = new GoogleAICacheManager("GOOGLE_API_KEY");
const fileManager = new GoogleAIFileManager("GOOGLE_API_KEY");

const uploadResult = await fileManager.uploadFile("path/to/a11.txt", {
  mimeType: "text/plain",
});

const cacheResult = await cacheManager.create({
  model: "models/gemini-1.5-flash",
  contents: [
    {
      role: "user",
      parts: [
        {
          fileData: {
            fileUri: uploadResult.file.uri,
            mimeType: uploadResult.file.mimeType,
          },
        },
      ],
    },
  ],
});

console.log(cacheResult);

const genAI = new GoogleGenerativeAI("GOOGLE_API_KEY");
const model = genAI.getGenerativeModelFromCachedContent(cacheResult);
const result = await model.generateContent(
  "Please summarize this transcript.",
);
console.log(result.response.text());

Dopo

Python

import requests
import pathlib
from google import genai
from google.genai import types

client = genai.Client()

# Check which models support caching.
for m in client.models.list():
  for action in m.supported_actions:
    if action == "createCachedContent":
      print(m.name)
      break

# Download file
response = requests.get(
    'https://storage.googleapis.com/generativeai-downloads/data/a11.txt')
pathlib.Path('a11.txt').write_text(response.text)

# Upload file
document = client.files.upload(file='a11.txt')

# Create cache
model='gemini-1.5-flash-001'
apollo_cache = client.caches.create(
      model=model,
      config={
          'contents': [document],
          'system_instruction': 'You are an expert at analyzing transcripts.',
      },
  )

# Generate response
response = client.models.generate_content(
    model=model,
    contents='Find a lighthearted moment from this transcript',
    config=types.GenerateContentConfig(
        cached_content=apollo_cache.name,
    )
)

JavaScript

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

const ai = new GoogleGenAI({ apiKey: "GOOGLE_API_KEY" });
const filePath = path.join(media, "a11.txt");
const document = await ai.files.upload({
  file: filePath,
  config: { mimeType: "text/plain" },
});
console.log("Uploaded file name:", document.name);
const modelName = "gemini-1.5-flash";

const contents = [
  createUserContent(createPartFromUri(document.uri, document.mimeType)),
];

const cache = await ai.caches.create({
  model: modelName,
  config: {
    contents: contents,
    systemInstruction: "You are an expert analyzing transcripts.",
  },
});
console.log("Cache created:", cache);

const response = await ai.models.generateContent({
  model: modelName,
  contents: "Please summarize this transcript",
  config: { cachedContent: cache.name },
});
console.log("Response text:", response.text);

Conteggio dei token

Conta il numero di token in una richiesta.

Prima

Python

import google.generativeai as genai

model = genai.GenerativeModel('gemini-1.5-flash')
response = model.count_tokens(
    'The quick brown fox jumps over the lazy dog.')

JavaScript

 import { GoogleGenerativeAI } from "@google/generative-ai";

 const genAI = new GoogleGenerativeAI("GOOGLE_API_KEY+);
 const model = genAI.getGenerativeModel({
   model: "gemini-1.5-flash",
 });

 // Count tokens in a prompt without calling text generation.
 const countResult = await model.countTokens(
   "The quick brown fox jumps over the lazy dog.",
 );

 console.log(countResult.totalTokens); // 11

 const generateResult = await model.generateContent(
   "The quick brown fox jumps over the lazy dog.",
 );

 // On the response for `generateContent`, use `usageMetadata`
 // to get separate input and output token counts
 // (`promptTokenCount` and `candidatesTokenCount`, respectively),
 // as well as the combined token count (`totalTokenCount`).
 console.log(generateResult.response.usageMetadata);
 // candidatesTokenCount and totalTokenCount depend on response, may vary
 // { promptTokenCount: 11, candidatesTokenCount: 124, totalTokenCount: 135 }

Dopo

Python

from google import genai

client = genai.Client()

response = client.models.count_tokens(
    model='gemini-2.0-flash',
    contents='The quick brown fox jumps over the lazy dog.',
)

JavaScript

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

const ai = new GoogleGenAI({ apiKey: "GOOGLE_API_KEY" });
const prompt = "The quick brown fox jumps over the lazy dog.";
const countTokensResponse = await ai.models.countTokens({
  model: "gemini-2.0-flash",
  contents: prompt,
});
console.log(countTokensResponse.totalTokens);

const generateResponse = await ai.models.generateContent({
  model: "gemini-2.0-flash",
  contents: prompt,
});
console.log(generateResponse.usageMetadata);

Genera immagini

Genera immagini:

Prima

Python

#pip install https://github.com/google-gemini/generative-ai-python@imagen
import google.generativeai as genai

imagen = genai.ImageGenerationModel(
    "imagen-3.0-generate-001")
gen_images = imagen.generate_images(
    prompt="Robot holding a red skateboard",
    number_of_images=1,
    safety_filter_level="block_low_and_above",
    person_generation="allow_adult",
    aspect_ratio="3:4",
)

Dopo

Python

from google import genai

client = genai.Client()

gen_images = client.models.generate_images(
    model='imagen-3.0-generate-001',
    prompt='Robot holding a red skateboard',
    config=types.GenerateImagesConfig(
        number_of_images= 1,
        safety_filter_level= "BLOCK_LOW_AND_ABOVE",
        person_generation= "ALLOW_ADULT",
        aspect_ratio= "3:4",
    )
)

for n, image in enumerate(gen_images.generated_images):
    pathlib.Path(f'{n}.png').write_bytes(
        image.image.image_bytes)

Incorporare contenuti

Generare incorporamenti di contenuti.

Prima

Python

import google.generativeai as genai

response = genai.embed_content(
  model='models/gemini-embedding-001',
  content='Hello world'
)

JavaScript

import { GoogleGenerativeAI } from "@google/generative-ai";

const genAI = new GoogleGenerativeAI("GOOGLE_API_KEY");
const model = genAI.getGenerativeModel({
  model: "gemini-embedding-001",
});

const result = await model.embedContent("Hello world!");

console.log(result.embedding);

Dopo

Python

from google import genai

client = genai.Client()

response = client.models.embed_content(
  model='gemini-embedding-001',
  contents='Hello world',
)

JavaScript

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

const ai = new GoogleGenAI({ apiKey: "GOOGLE_API_KEY" });
const text = "Hello World!";
const result = await ai.models.embedContent({
  model: "gemini-embedding-001",
  contents: text,
  config: { outputDimensionality: 10 },
});
console.log(result.embeddings);

Ottimizza un modello

Crea e utilizza un modello ottimizzato.

Il nuovo SDK semplifica la regolazione con client.tunings.tune, che avvia il job di regolazione ed esegue il polling finché non viene completato.

Prima

Python

import google.generativeai as genai
import random

# create tuning model
train_data = {}
for i in range(1, 6):
  key = f'input {i}'
  value = f'output {i}'
  train_data[key] = value

name = f'generate-num-{random.randint(0,10000)}'
operation = genai.create_tuned_model(
    source_model='models/gemini-1.5-flash-001-tuning',
    training_data=train_data,
    id = name,
    epoch_count = 5,
    batch_size=4,
    learning_rate=0.001,
)
# wait for tuning complete
tuningProgress = operation.result()

# generate content with the tuned model
model = genai.GenerativeModel(model_name=f'tunedModels/{name}')
response = model.generate_content('55')

Dopo

Python

from google import genai
from google.genai import types

client = genai.Client()

# Check which models are available for tuning.
for m in client.models.list():
  for action in m.supported_actions:
    if action == "createTunedModel":
      print(m.name)
      break

# create tuning model
training_dataset=types.TuningDataset(
        examples=[
            types.TuningExample(
                text_input=f'input {i}',
                output=f'output {i}',
            )
            for i in range(5)
        ],
    )
tuning_job = client.tunings.tune(
    base_model='models/gemini-1.5-flash-001-tuning',
    training_dataset=training_dataset,
    config=types.CreateTuningJobConfig(
        epoch_count= 5,
        batch_size=4,
        learning_rate=0.001,
        tuned_model_display_name="test tuned model"
    )
)

# generate content with the tuned model
response = client.models.generate_content(
    model=tuning_job.tuned_model.model,
    contents='55',
)