Google Gemini API Setup

Self-Hosting Documentation Access Granted

LlamaCloud supports Google Gemini API for direct access to Google’s AI models with simple API key authentication. This provides a straightforward alternative to Google Vertex AI when you don’t need enterprise Google Cloud Platform features.

Prerequisites

A valid Google account
Google Gemini API key from Google AI Studio
Access and quota for supported models:
- Gemini 1.5 Pro
- Gemini 1.5 Flash
- Gemini 2.0 Flash

Environment Variables

The Google Gemini API integration uses a single environment variable:

GOOGLE_GEMINI_API_KEY - Your Google Gemini API key (required)

Configuration

Follow these steps to configure Google Gemini API integration:

Step 1: Get Google Gemini API Key

Get your API key from Google AI Studio.

Step 2: Create Kubernetes Secret

Create a secret with your Google Gemini API key:

apiVersion: v1
kind: Secret
metadata:
  name: gemini-credentials
type: Opaque
stringData:
  GOOGLE_GEMINI_API_KEY: "your-api-key-here"

Apply the secret:

kubectl apply -f gemini-secret.yaml

Step 3: Configure Helm Values

Reference the secret in your Helm configuration:

# External Secret
config:
  llms:
    gemini:
      secret: "gemini-credentials"

# or direct configuration
config:
  llms:
    gemini:
      apiKey: "your-api-key-here"

Verification

After configuration, verify your Google Gemini integration:

Verify in Admin UI: Check available Google models in LlamaCloud admin interface
Test functionality: Upload a document to confirm Gemini models are working

Troubleshooting

Common Issues

API Key Invalid

Error: API key not valid

Solution:

Ensure your API key is correctly set and hasn’t expired
Verify the key is from Google AI Studio
Check that the API key has proper permissions

Quota Exceeded

Error: Quota exceeded

Solution:

Check your Google AI Studio quotas and usage limits
Consider upgrading your plan or requesting quota increases
Monitor API usage to avoid rate limiting

Model Access Issues

Error: Model not found or access denied

Solution:

Verify the model is available in your region
Check if you have access to the specific model
Ensure your API key has model access permissions

Debug Steps

Test Gemini API directly:

curl -H "Content-Type: application/json" \
     -H "x-goog-api-key: YOUR_API_KEY" \
     -d '{"contents":[{"parts":[{"text":"Hello"}]}]}' \
     "https://generativelanguage.googleapis.com/v1beta/models/gemini-pro:generateContent"

Verify secret mounting:

kubectl get secret gemini-credentials -o yaml
kubectl describe pod <pod-name> | grep -A 20 Environment

Check network connectivity: Ensure your cluster can reach generativelanguage.googleapis.com