A API Gemini pode gerar saída de texto de várias entradas, incluindo texto, imagens, vídeo e áudio, usando os modelos Gemini.
Confira um exemplo básico que usa uma única entrada de texto:
from google import genai
client = genai.Client(api_key="GEMINI_API_KEY")
response = client.models.generate_content(
model="gemini-2.0-flash",
contents=["How does AI work?"]
)
print(response.text)
import { GoogleGenAI } from "@google/genai";
const ai = new GoogleGenAI({ apiKey: "GEMINI_API_KEY" });
async function main() {
const response = await ai.models.generateContent({
model: "gemini-2.0-flash",
contents: "How does AI work?",
});
console.log(response.text);
}
await main();
package main
import (
"context"
"fmt"
"os"
"google.golang.org/genai"
)
func main() {
ctx := context.Background()
client, _ := genai.NewClient(ctx, &genai.ClientConfig{
APIKey: os.Getenv("GEMINI_API_KEY"),
Backend: genai.BackendGeminiAPI,
})
result, _ := client.Models.GenerateContent(
ctx,
"gemini-2.0-flash",
genai.Text("Explain how AI works in a few words"),
nil,
)
fmt.Println(result.Text())
}
curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=$GEMINI_API_KEY" \
-H 'Content-Type: application/json' \
-X POST \
-d '{
"contents": [
{
"parts": [
{
"text": "How does AI work?"
}
]
}
]
}'
// See https://developers.google.com/apps-script/guides/properties
// for instructions on how to set the API key.
const apiKey = PropertiesService.getScriptProperties().getProperty('GEMINI_API_KEY');
function main() {
const payload = {
contents: [
{
parts: [
{ text: 'How AI does work?' },
],
},
],
};
const url = `https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=${apiKey}`;
const options = {
method: 'POST',
contentType: 'application/json',
payload: JSON.stringify(payload)
};
const response = UrlFetchApp.fetch(url, options);
const data = JSON.parse(response);
const content = data['candidates'][0]['content']['parts'][0]['text'];
console.log(content);
}
Instruções e configuração do sistema
É possível orientar o comportamento dos modelos do Gemini com instruções do sistema. Para fazer isso,
transmita um objeto GenerateContentConfig
.
from google import genai
from google.genai import types
client = genai.Client(api_key="GEMINI_API_KEY")
response = client.models.generate_content(
model="gemini-2.0-flash",
config=types.GenerateContentConfig(
system_instruction="You are a cat. Your name is Neko."),
contents="Hello there"
)
print(response.text)
import { GoogleGenAI } from "@google/genai";
const ai = new GoogleGenAI({ apiKey: "GEMINI_API_KEY" });
async function main() {
const response = await ai.models.generateContent({
model: "gemini-2.0-flash",
contents: "Hello there",
config: {
systemInstruction: "You are a cat. Your name is Neko.",
},
});
console.log(response.text);
}
await main();
package main
import (
"context"
"fmt"
"os"
"google.golang.org/genai"
)
func main() {
ctx := context.Background()
client, _ := genai.NewClient(ctx, &genai.ClientConfig{
APIKey: os.Getenv("GEMINI_API_KEY"),
Backend: genai.BackendGeminiAPI,
})
config := &genai.GenerateContentConfig{
SystemInstruction: genai.NewContentFromText("You are a cat. Your name is Neko.", genai.RoleUser),
}
result, _ := client.Models.GenerateContent(
ctx,
"gemini-2.0-flash",
genai.Text("Hello there"),
config,
)
fmt.Println(result.Text())
}
curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=$GEMINI_API_KEY" \
-H 'Content-Type: application/json' \
-d '{
"system_instruction": {
"parts": [
{
"text": "You are a cat. Your name is Neko."
}
]
},
"contents": [
{
"parts": [
{
"text": "Hello there"
}
]
}
]
}'
// See https://developers.google.com/apps-script/guides/properties
// for instructions on how to set the API key.
const apiKey = PropertiesService.getScriptProperties().getProperty('GEMINI_API_KEY');
function main() {
const systemInstruction = {
parts: [{
text: 'You are a cat. Your name is Neko.'
}]
};
const payload = {
systemInstruction,
contents: [
{
parts: [
{ text: 'Hello there' },
],
},
],
};
const url = `https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=${apiKey}`;
const options = {
method: 'POST',
contentType: 'application/json',
payload: JSON.stringify(payload)
};
const response = UrlFetchApp.fetch(url, options);
const data = JSON.parse(response);
const content = data['candidates'][0]['content']['parts'][0]['text'];
console.log(content);
}
O objeto GenerateContentConfig
também permite substituir parâmetros de geração padrão, como
temperature.
from google import genai
from google.genai import types
client = genai.Client(api_key="GEMINI_API_KEY")
response = client.models.generate_content(
model="gemini-2.0-flash",
contents=["Explain how AI works"],
config=types.GenerateContentConfig(
max_output_tokens=500,
temperature=0.1
)
)
print(response.text)
import { GoogleGenAI } from "@google/genai";
const ai = new GoogleGenAI({ apiKey: "GEMINI_API_KEY" });
async function main() {
const response = await ai.models.generateContent({
model: "gemini-2.0-flash",
contents: "Explain how AI works",
config: {
maxOutputTokens: 500,
temperature: 0.1,
},
});
console.log(response.text);
}
await main();
package main
import (
"context"
"fmt"
"os"
"google.golang.org/genai"
)
func main() {
ctx := context.Background()
client, _ := genai.NewClient(ctx, &genai.ClientConfig{
APIKey: os.Getenv("GEMINI_API_KEY"),
Backend: genai.BackendGeminiAPI,
})
temp := float32(0.9)
topP := float32(0.5)
topK := float32(20.0)
maxOutputTokens := int32(100)
config := &genai.GenerateContentConfig{
Temperature: &temp,
TopP: &topP,
TopK: &topK,
MaxOutputTokens: maxOutputTokens,
ResponseMIMEType: "application/json",
}
result, _ := client.Models.GenerateContent(
ctx,
"gemini-2.0-flash",
genai.Text("What is the average size of a swallow?"),
config,
)
fmt.Println(result.Text())
}
curl https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=$GEMINI_API_KEY \
-H 'Content-Type: application/json' \
-X POST \
-d '{
"contents": [
{
"parts": [
{
"text": "Explain how AI works"
}
]
}
],
"generationConfig": {
"stopSequences": [
"Title"
],
"temperature": 1.0,
"maxOutputTokens": 800,
"topP": 0.8,
"topK": 10
}
}'
// See https://developers.google.com/apps-script/guides/properties
// for instructions on how to set the API key.
const apiKey = PropertiesService.getScriptProperties().getProperty('GEMINI_API_KEY');
function main() {
const generationConfig = {
temperature: 1,
topP: 0.95,
topK: 40,
maxOutputTokens: 8192,
responseMimeType: 'text/plain',
};
const payload = {
generationConfig,
contents: [
{
parts: [
{ text: 'Explain how AI works in a few words' },
],
},
],
};
const url = `https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=${apiKey}`;
const options = {
method: 'POST',
contentType: 'application/json',
payload: JSON.stringify(payload)
};
const response = UrlFetchApp.fetch(url, options);
const data = JSON.parse(response);
const content = data['candidates'][0]['content']['parts'][0]['text'];
console.log(content);
}
Consulte o GenerateContentConfig
na nossa referência da API para conferir uma lista completa de parâmetros configuráveis e as
descrições deles.
Entradas multimodais
A API Gemini oferece suporte a entradas multimodais, permitindo combinar texto com arquivos de mídia. O exemplo a seguir demonstra como fornecer uma imagem:
from PIL import Image
from google import genai
client = genai.Client(api_key="GEMINI_API_KEY")
image = Image.open("/path/to/organ.png")
response = client.models.generate_content(
model="gemini-2.0-flash",
contents=[image, "Tell me about this instrument"]
)
print(response.text)
import {
GoogleGenAI,
createUserContent,
createPartFromUri,
} from "@google/genai";
const ai = new GoogleGenAI({ apiKey: "GEMINI_API_KEY" });
async function main() {
const image = await ai.files.upload({
file: "/path/to/organ.png",
});
const response = await ai.models.generateContent({
model: "gemini-2.0-flash",
contents: [
createUserContent([
"Tell me about this instrument",
createPartFromUri(image.uri, image.mimeType),
]),
],
});
console.log(response.text);
}
await main();
package main
import (
"context"
"fmt"
"os"
"google.golang.org/genai"
)
func main() {
ctx := context.Background()
client, _ := genai.NewClient(ctx, &genai.ClientConfig{
APIKey: os.Getenv("GEMINI_API_KEY"),
Backend: genai.BackendGeminiAPI,
})
imagePath := "/path/to/organ.jpg"
imgData, _ := os.ReadFile(imagePath)
parts := []*genai.Part{
genai.NewPartFromText("Tell me about this instrument"),
&genai.Part{
InlineData: &genai.Blob{
MIMEType: "image/jpeg",
Data: imgData,
},
},
}
contents := []*genai.Content{
genai.NewContentFromParts(parts, genai.RoleUser),
}
result, _ := client.Models.GenerateContent(
ctx,
"gemini-2.0-flash",
contents,
nil,
)
fmt.Println(result.Text())
}
# Use a temporary file to hold the base64 encoded image data
TEMP_B64=$(mktemp)
trap 'rm -f "$TEMP_B64"' EXIT
base64 $B64FLAGS $IMG_PATH > "$TEMP_B64"
# Use a temporary file to hold the JSON payload
TEMP_JSON=$(mktemp)
trap 'rm -f "$TEMP_JSON"' EXIT
cat > "$TEMP_JSON" << EOF
{
"contents": [
{
"parts": [
{
"text": "Tell me about this instrument"
},
{
"inline_data": {
"mime_type": "image/jpeg",
"data": "$(cat "$TEMP_B64")"
}
}
]
}
]
}
EOF
curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=$GEMINI_API_KEY" \
-H 'Content-Type: application/json' \
-X POST \
-d "@$TEMP_JSON"
// See https://developers.google.com/apps-script/guides/properties
// for instructions on how to set the API key.
const apiKey = PropertiesService.getScriptProperties().getProperty('GEMINI_API_KEY');
function main() {
const imageUrl = 'http://image/url';
const image = getImageData(imageUrl);
const payload = {
contents: [
{
parts: [
{ image },
{ text: 'Tell me about this instrument' },
],
},
],
};
const url = `https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=${apiKey}`;
const options = {
method: 'POST',
contentType: 'application/json',
payload: JSON.stringify(payload)
};
const response = UrlFetchApp.fetch(url, options);
const data = JSON.parse(response);
const content = data['candidates'][0]['content']['parts'][0]['text'];
console.log(content);
}
function getImageData(url) {
const blob = UrlFetchApp.fetch(url).getBlob();
return {
mimeType: blob.getContentType(),
data: Utilities.base64Encode(blob.getBytes())
};
}
Para métodos alternativos de envio de imagens e processamento de imagens mais avançado, consulte nosso guia de compreensão de imagens. A API também oferece suporte a entradas e compreensão de documentos, vídeos e áudio.
Respostas de streaming
Por padrão, o modelo retorna uma resposta somente depois que todo o processo de geração é concluído.
Para interações mais fluidas, use a transmissão para receber instâncias de GenerateContentResponse
de forma incremental
conforme são geradas.
from google import genai
client = genai.Client(api_key="GEMINI_API_KEY")
response = client.models.generate_content_stream(
model="gemini-2.0-flash",
contents=["Explain how AI works"]
)
for chunk in response:
print(chunk.text, end="")
import { GoogleGenAI } from "@google/genai";
const ai = new GoogleGenAI({ apiKey: "GEMINI_API_KEY" });
async function main() {
const response = await ai.models.generateContentStream({
model: "gemini-2.0-flash",
contents: "Explain how AI works",
});
for await (const chunk of response) {
console.log(chunk.text);
}
}
await main();
package main
import (
"context"
"fmt"
"os"
"google.golang.org/genai"
)
func main() {
ctx := context.Background()
client, _ := genai.NewClient(ctx, &genai.ClientConfig{
APIKey: os.Getenv("GEMINI_API_KEY"),
Backend: genai.BackendGeminiAPI,
})
stream := client.Models.GenerateContentStream(
ctx,
"gemini-2.0-flash",
genai.Text("Write a story about a magic backpack."),
nil,
)
for chunk, _ := range stream {
part := chunk.Candidates[0].Content.Parts[0]
fmt.Print(part.Text)
}
}
curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:streamGenerateContent?alt=sse&key=${GEMINI_API_KEY}" \
-H 'Content-Type: application/json' \
--no-buffer \
-d '{
"contents": [
{
"parts": [
{
"text": "Explain how AI works"
}
]
}
]
}'
// See https://developers.google.com/apps-script/guides/properties
// for instructions on how to set the API key.
const apiKey = PropertiesService.getScriptProperties().getProperty('GEMINI_API_KEY');
function main() {
const payload = {
contents: [
{
parts: [
{ text: 'Explain how AI works' },
],
},
],
};
const url = `https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:streamGenerateContent?key=${apiKey}`;
const options = {
method: 'POST',
contentType: 'application/json',
payload: JSON.stringify(payload)
};
const response = UrlFetchApp.fetch(url, options);
const data = JSON.parse(response);
const content = data['candidates'][0]['content']['parts'][0]['text'];
console.log(content);
}
Conversas com vários turnos (chat)
Nossos SDKs oferecem funcionalidade para coletar várias rodadas de comandos e respostas em uma conversa, facilitando o acompanhamento do histórico da conversa.
from google import genai
client = genai.Client(api_key="GEMINI_API_KEY")
chat = client.chats.create(model="gemini-2.0-flash")
response = chat.send_message("I have 2 dogs in my house.")
print(response.text)
response = chat.send_message("How many paws are in my house?")
print(response.text)
for message in chat.get_history():
print(f'role - {message.role}',end=": ")
print(message.parts[0].text)
import { GoogleGenAI } from "@google/genai";
const ai = new GoogleGenAI({ apiKey: "GEMINI_API_KEY" });
async function main() {
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);
}
await main();
package main
import (
"context"
"fmt"
"os"
"google.golang.org/genai"
)
func main() {
ctx := context.Background()
client, _ := genai.NewClient(ctx, &genai.ClientConfig{
APIKey: os.Getenv("GEMINI_API_KEY"),
Backend: genai.BackendGeminiAPI,
})
history := []*genai.Content{
genai.NewContentFromText("Hi nice to meet you! I have 2 dogs in my house.", genai.RoleUser),
genai.NewContentFromText("Great to meet you. What would you like to know?", genai.RoleModel),
}
chat, _ := client.Chats.Create(ctx, "gemini-2.0-flash", nil, history)
res, _ := chat.SendMessage(ctx, genai.Part{Text: "How many paws are in my house?"})
if len(res.Candidates) > 0 {
fmt.Println(res.Candidates[0].Content.Parts[0].Text)
}
}
curl https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=$GEMINI_API_KEY \
-H 'Content-Type: application/json' \
-X POST \
-d '{
"contents": [
{
"role": "user",
"parts": [
{
"text": "Hello"
}
]
},
{
"role": "model",
"parts": [
{
"text": "Great to meet you. What would you like to know?"
}
]
},
{
"role": "user",
"parts": [
{
"text": "I have two dogs in my house. How many paws are in my house?"
}
]
}
]
}'
// See https://developers.google.com/apps-script/guides/properties
// for instructions on how to set the API key.
const apiKey = PropertiesService.getScriptProperties().getProperty('GEMINI_API_KEY');
function main() {
const payload = {
contents: [
{
role: 'user',
parts: [
{ text: 'Hello' },
],
},
{
role: 'model',
parts: [
{ text: 'Great to meet you. What would you like to know?' },
],
},
{
role: 'user',
parts: [
{ text: 'I have two dogs in my house. How many paws are in my house?' },
],
},
],
};
const url = `https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:generateContent?key=${apiKey}`;
const options = {
method: 'POST',
contentType: 'application/json',
payload: JSON.stringify(payload)
};
const response = UrlFetchApp.fetch(url, options);
const data = JSON.parse(response);
const content = data['candidates'][0]['content']['parts'][0]['text'];
console.log(content);
}
O streaming também pode ser usado para conversas com vários turnos.
from google import genai
client = genai.Client(api_key="GEMINI_API_KEY")
chat = client.chats.create(model="gemini-2.0-flash")
response = chat.send_message_stream("I have 2 dogs in my house.")
for chunk in response:
print(chunk.text, end="")
response = chat.send_message_stream("How many paws are in my house?")
for chunk in response:
print(chunk.text, end="")
for message in chat.get_history():
print(f'role - {message.role}', end=": ")
print(message.parts[0].text)
import { GoogleGenAI } from "@google/genai";
const ai = new GoogleGenAI({ apiKey: "GEMINI_API_KEY" });
async function main() {
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 stream1 = await chat.sendMessageStream({
message: "I have 2 dogs in my house.",
});
for await (const chunk of stream1) {
console.log(chunk.text);
console.log("_".repeat(80));
}
const stream2 = await chat.sendMessageStream({
message: "How many paws are in my house?",
});
for await (const chunk of stream2) {
console.log(chunk.text);
console.log("_".repeat(80));
}
}
await main();
package main
import (
"context"
"fmt"
"os"
"google.golang.org/genai"
)
func main() {
ctx := context.Background()
client, _ := genai.NewClient(ctx, &genai.ClientConfig{
APIKey: os.Getenv("GEMINI_API_KEY"),
Backend: genai.BackendGeminiAPI,
})
history := []*genai.Content{
genai.NewContentFromText("Hi nice to meet you! I have 2 dogs in my house.", genai.RoleUser),
genai.NewContentFromText("Great to meet you. What would you like to know?", genai.RoleModel),
}
chat, _ := client.Chats.Create(ctx, "gemini-2.0-flash", nil, history)
stream := chat.SendMessageStream(ctx, genai.Part{Text: "How many paws are in my house?"})
for chunk, _ := range stream {
part := chunk.Candidates[0].Content.Parts[0]
fmt.Print(part.Text)
}
}
curl https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:streamGenerateContent?alt=sse&key=$GEMINI_API_KEY \
-H 'Content-Type: application/json' \
-X POST \
-d '{
"contents": [
{
"role": "user",
"parts": [
{
"text": "Hello"
}
]
},
{
"role": "model",
"parts": [
{
"text": "Great to meet you. What would you like to know?"
}
]
},
{
"role": "user",
"parts": [
{
"text": "I have two dogs in my house. How many paws are in my house?"
}
]
}
]
}'
// See https://developers.google.com/apps-script/guides/properties
// for instructions on how to set the API key.
const apiKey = PropertiesService.getScriptProperties().getProperty('GEMINI_API_KEY');
function main() {
const payload = {
contents: [
{
role: 'user',
parts: [
{ text: 'Hello' },
],
},
{
role: 'model',
parts: [
{ text: 'Great to meet you. What would you like to know?' },
],
},
{
role: 'user',
parts: [
{ text: 'I have two dogs in my house. How many paws are in my house?' },
],
},
],
};
const url = `https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash:streamGenerateContent?key=${apiKey}`;
const options = {
method: 'POST',
contentType: 'application/json',
payload: JSON.stringify(payload)
};
const response = UrlFetchApp.fetch(url, options);
const data = JSON.parse(response);
const content = data['candidates'][0]['content']['parts'][0]['text'];
console.log(content);
}
Modelos compatíveis
Todos os modelos da família Gemini oferecem suporte à geração de texto. Para saber mais sobre os modelos e os recursos deles, acesse a página Modelos.
Práticas recomendadas
Dicas de comandos
Para a geração de texto básica, um comando zero-shot muitas vezes é suficiente sem precisar de exemplos, instruções do sistema ou formatação específica.
Para saídas mais personalizadas:
- Use Instruções do sistema para orientar o modelo.
- Forneça alguns exemplos de entradas e saídas para orientar o modelo. Isso é chamado de comando de poucos disparos.
- Considere ajustar para casos de uso avançados.
Consulte nosso guia de engenharia de comandos para mais dicas.
Resposta estruturada
Em alguns casos, talvez seja necessário uma saída estruturada, como JSON. Consulte nosso guia de saída estruturada para saber como fazer isso.
A seguir
- Confira o Guia de início da API Gemini no Colab.
- Conheça os recursos de compreensão de imagem, vídeo, áudio e documento do Gemini.
- Saiba mais sobre as estratégias de comando de arquivos multimodais.