Google Cloud APIs Docs

Appearance

Client Libraries Guide ​

✅ Fresh -- Updated March 2026 from official Google Cloud documentation.

Overview ​

Google provides idiomatic client libraries for all major programming languages. These libraries handle authentication, retries, pagination, and serialization automatically.

Supported Languages ​

LanguagePackage ManagerInstall Pattern
Pythonpippip install google-cloud-SERVICE
Node.jsnpmnpm install @google-cloud/SERVICE
JavaMaven/Gradlecom.google.cloud:google-cloud-SERVICE
Gogo modulesgo get cloud.google.com/go/SERVICE
C#NuGetdotnet add package Google.Cloud.SERVICE.V1
PHPComposercomposer require google/cloud-SERVICE
Rubygemgem install google-cloud-SERVICE

Installation ​

Python ​

bash
# Install a specific service library
pip install google-cloud-storage
pip install google-cloud-bigquery
pip install google-cloud-pubsub

# Install with extras (e.g., pandas integration for BigQuery)
pip install google-cloud-bigquery[pandas]

Node.js ​

bash
npm install @google-cloud/storage
npm install @google-cloud/bigquery
npm install @google-cloud/pubsub

Java (Maven) ​

xml
<dependency>
  <groupId>com.google.cloud</groupId>
  <artifactId>google-cloud-storage</artifactId>
  <version>2.30.0</version>
</dependency>

Go ​

bash
go get cloud.google.com/go/storage
go get cloud.google.com/go/bigquery
go get cloud.google.com/go/pubsub

Authentication in Client Libraries ​

All client libraries use Application Default Credentials (ADC) automatically:

Python ​

python
from google.cloud import storage

# ADC is used automatically
client = storage.Client()

# Or specify a project explicitly
client = storage.Client(project='my-project')

# Or use a specific credentials file
from google.oauth2 import service_account
creds = service_account.Credentials.from_service_account_file('key.json')
client = storage.Client(credentials=creds, project='my-project')

Node.js ​

javascript
const { Storage } = require('@google-cloud/storage');

// ADC is used automatically
const storage = new Storage();

// Or specify project and key file
const storage = new Storage({
  projectId: 'my-project',
  keyFilename: '/path/to/key.json'
});

Go ​

go
import "cloud.google.com/go/storage"

// ADC is used automatically
client, err := storage.NewClient(ctx)

// Or specify credentials
client, err := storage.NewClient(ctx,
    option.WithCredentialsFile("/path/to/key.json"))

Common Operations ​

Cloud Storage ​

python
from google.cloud import storage

client = storage.Client()

# List buckets
for bucket in client.list_buckets():
    print(bucket.name)

# Upload a file
bucket = client.bucket('my-bucket')
blob = bucket.blob('path/to/file.txt')
blob.upload_from_filename('/local/file.txt')

# Download a file
blob.download_to_filename('/local/download.txt')

# Delete a file
blob.delete()

BigQuery ​

python
from google.cloud import bigquery

client = bigquery.Client()

# Run a query
query = "SELECT name, count FROM `my-project.my_dataset.my_table` LIMIT 10"
results = client.query(query)

for row in results:
    print(f"{row.name}: {row.count}")

Pub/Sub ​

python
from google.cloud import pubsub_v1

# Publish a message
publisher = pubsub_v1.PublisherClient()
topic_path = publisher.topic_path('my-project', 'my-topic')
future = publisher.publish(topic_path, b'Hello, World!')
print(f"Published message ID: {future.result()}")

# Subscribe to messages
subscriber = pubsub_v1.SubscriberClient()
subscription_path = subscriber.subscription_path('my-project', 'my-sub')

def callback(message):
    print(f"Received: {message.data}")
    message.ack()

streaming_pull = subscriber.subscribe(subscription_path, callback=callback)

Automatic Retry ​

Client libraries implement automatic retry with exponential backoff for transient errors:

Customizing Retry Behavior ​

Python:

python
from google.api_core.retry import Retry

# Custom retry with different timeout
custom_retry = Retry(
    initial=1.0,         # Initial backoff in seconds
    maximum=60.0,        # Max backoff
    multiplier=2.0,      # Backoff multiplier
    deadline=300.0,      # Total deadline in seconds
)

client.get_bucket('my-bucket', retry=custom_retry)

Node.js:

javascript
const { Storage } = require('@google-cloud/storage');

const storage = new Storage({
  retryOptions: {
    autoRetry: true,
    maxRetries: 5,
    retryDelayMultiplier: 2,
    totalTimeout: 300,
  }
});

Pagination ​

Client libraries handle pagination automatically:

python
from google.cloud import storage

client = storage.Client()

# Automatic iteration across all pages
for blob in client.list_blobs('my-bucket'):
    print(blob.name)

# Manual pagination
blobs = client.list_blobs('my-bucket', max_results=100)
page = next(blobs.pages)
for blob in page:
    print(blob.name)
print(f"Next page token: {blobs.next_page_token}")

Error Handling ​

python
from google.api_core.exceptions import NotFound, PermissionDenied, TooManyRequests

try:
    bucket = client.get_bucket('nonexistent-bucket')
except NotFound:
    print("Bucket does not exist")
except PermissionDenied:
    print("No permission to access bucket")
except TooManyRequests:
    print("Rate limited -- slow down")
except Exception as e:
    print(f"Unexpected error: {e}")

Library Architecture ​

Best Practices ​

  1. Always use client libraries instead of raw HTTP calls
  2. Use ADC for authentication (no hardcoded keys)
  3. Handle specific exceptions (NotFound, PermissionDenied) not just generic Exception
  4. Close clients when done to release resources (client.close())
  5. Use async clients for high-throughput applications
  6. Keep libraries updated for security patches and performance improvements
bash
# Check for updates
pip list --outdated | grep google-cloud
npm outdated | grep @google-cloud

See Also ​