How to beginner · 3 min read

How to set up Google AI Studio

Q: How to set up Google AI Studio

To set up Google AI Studio, create a Google Cloud project, enable the Gemini API, and configure authentication with a service account key. Use the google-cloud-aiplatform Python SDK to interact with Gemini models programmatically.

Quick answer

To set up Google AI Studio, create a Google Cloud project, enable the Gemini API, and configure authentication with a service account key. Use the google-cloud-aiplatform Python SDK to interact with Gemini models programmatically.

PREREQUISITES

Python 3.8+
Google Cloud account
Google Cloud SDK installed
Enable Gemini API in Google Cloud Console
pip install google-cloud-aiplatform

Setup

Install the Google Cloud SDK and the AI Platform Python client library. Set up authentication by creating a service account with the necessary permissions and download its JSON key file. Set the GOOGLE_APPLICATION_CREDENTIALS environment variable to point to this key file.

bash

# Install Google Cloud SDK (if not installed)
curl https://sdk.cloud.google.com | bash
exec -l $SHELL

# Install AI Platform client library
pip install google-cloud-aiplatform

# Set environment variable for authentication
export GOOGLE_APPLICATION_CREDENTIALS="/path/to/your/service-account-key.json"

Step by step

This example shows how to initialize the AI Platform client and send a text prompt to a Gemini model using Python.

python

from google.cloud import aiplatform
import os

# Initialize AI Platform client
client = aiplatform.gapic.PredictionServiceClient()

# Your Google Cloud project and location
project = "your-project-id"
location = "us-central1"
endpoint_id = "your-gemini-endpoint-id"  # Replace with your deployed Gemini endpoint

endpoint = client.endpoint_path(project=project, location=location, endpoint=endpoint_id)

# Prepare the prediction request
instances = [{"content": "Hello, how can I use Google AI Studio with Gemini?"}]
parameters = {}

response = client.predict(endpoint=endpoint, instances=instances, parameters=parameters)

print("Prediction response:", response.predictions)

output

Prediction response: ['Your Gemini model response here']

Common variations

You can use asynchronous calls with client.predict_async() for large workloads. Change the location to match your deployment region. Use different Gemini models by specifying their endpoint IDs. For streaming or chat-based interactions, use the appropriate client methods provided by the AI Platform SDK.

python

from google.cloud import aiplatform
import asyncio

async def async_predict():
    client = aiplatform.gapic.PredictionServiceAsyncClient()
    project = "your-project-id"
    location = "us-central1"
    endpoint_id = "your-gemini-endpoint-id"
    endpoint = client.endpoint_path(project=project, location=location, endpoint=endpoint_id)

    instances = [{"content": "Async call to Gemini model."}]
    parameters = {}

    response = await client.predict(endpoint=endpoint, instances=instances, parameters=parameters)
    print("Async prediction response:", response.predictions)

asyncio.run(async_predict())

output

Async prediction response: ['Your Gemini model async response here']

Troubleshooting

If you get PermissionDenied, verify your service account has the Vertex AI User role.
If endpoint not found, confirm the endpoint_id is correct and the model is deployed.
For authentication errors, ensure GOOGLE_APPLICATION_CREDENTIALS points to a valid JSON key file.

✅

Key Takeaways

Use the Google Cloud Console to enable Gemini API and create service accounts.
Set the environment variable GOOGLE_APPLICATION_CREDENTIALS for authentication.
Use the google-cloud-aiplatform Python SDK to call Gemini endpoints.
Async prediction calls improve performance for large or streaming workloads.
Check IAM roles and endpoint IDs carefully to avoid common errors.

Verified 2026-04 · gemini-1.5-pro, gemini-1.5-flash, gemini-2.0-flash

Verify ↗