How to beginner · 3 min read

How to authenticate with Vertex AI

Q: How to authenticate with Vertex AI

Authenticate with Vertex AI by setting up Google Cloud Application Default Credentials using gcloud auth application-default login or by providing a service account key JSON file via the GOOGLE_APPLICATION_CREDENTIALS environment variable. Then initialize the vertexai SDK in Python, which automatically uses these credentials for API calls.

Quick answer

Authenticate with Vertex AI by setting up Google Cloud Application Default Credentials using gcloud auth application-default login or by providing a service account key JSON file via the GOOGLE_APPLICATION_CREDENTIALS environment variable. Then initialize the vertexai SDK in Python, which automatically uses these credentials for API calls.

PREREQUISITES

Python 3.8+
Google Cloud SDK installed
Google Cloud project with Vertex AI enabled
Service account with Vertex AI permissions or user credentials
pip install vertexai

Setup

Install the vertexai Python SDK and configure authentication by either logging in with your user account or setting a service account key.

Install the SDK: pip install vertexai
Authenticate with user credentials: gcloud auth application-default login
Or set the environment variable for service account key JSON: export GOOGLE_APPLICATION_CREDENTIALS="/path/to/key.json"

bash

pip install vertexai

Step by step

This example shows how to initialize the vertexai SDK and generate a text completion using the gemini-2.0-flash model. The SDK automatically uses your authenticated credentials.

python

import os
import vertexai
from vertexai.generative_models import GenerativeModel

# Initialize Vertex AI with your Google Cloud project and location
vertexai.init(project=os.environ["GOOGLE_CLOUD_PROJECT"], location="us-central1")

# Load the Gemini model
model = GenerativeModel("gemini-2.0-flash")

# Generate content
response = model.generate_content("Explain quantum computing in simple terms.")
print(response.text)

output

Quantum computing is a type of computing that uses quantum bits, or qubits, which can represent both 0 and 1 at the same time, allowing computers to solve certain problems much faster than traditional computers.

Common variations

You can authenticate asynchronously or use different models by changing the model parameter. The vertexai SDK supports streaming responses by setting stream=True in generate_content.

python

import asyncio
import os
import vertexai
from vertexai.generative_models import GenerativeModel

async def async_generate():
    vertexai.init(project=os.environ["GOOGLE_CLOUD_PROJECT"], location="us-central1")
    model = GenerativeModel("gemini-2.5-pro")
    response = await model.generate_content("Summarize the benefits of AI.", stream=True)
    async for chunk in response:
        print(chunk.text, end="", flush=True)

asyncio.run(async_generate())

output

AI improves efficiency, automates tasks, enables new insights through data analysis, and enhances decision-making across industries.

Troubleshooting

If you see PermissionDenied errors, verify your service account has the Vertex AI User role.
If authentication fails, ensure GOOGLE_APPLICATION_CREDENTIALS points to a valid JSON key file or run gcloud auth application-default login.
Check your GOOGLE_CLOUD_PROJECT environment variable is set correctly.

✅

Key Takeaways

Use Application Default Credentials or service account keys to authenticate with Vertex AI.
Initialize the vertexai SDK with your project and location to make authenticated API calls.
For async or streaming, use generate_content with stream=True.
Ensure your Google Cloud project and roles are correctly configured to avoid permission errors.

Verified 2026-04 · gemini-2.0-flash, gemini-2.5-pro

Verify ↗