Google Vertex AI Piece

Integrate your automation workflows with Google Vertex AI to leverage powerful AI models including Gemini and other advanced language models.

Overview

The Google Vertex AI piece enables you to:

• Generate content and responses using Gemini and other Vertex AI models
• Make custom API requests to Vertex AI endpoints
• Build intelligent automation workflows with AI capabilities

Authentication

This piece uses Service Account Authentication for secure access to Google Vertex AI APIs.

How to Set Up

1. Create a Google Cloud Project

   • Go to Google Cloud Console
   • Create a new project or select an existing one

2. Enable Vertex AI API

   • Navigate to "APIs & Services" > "Enabled APIs & services"
   • Click "Enable APIs and Services"
   • Search for "Vertex AI API"
   • Click on it and press "Enable"

3. Create a Service Account

   • Go to "APIs & Services" > "Credentials"
   • Click "Create Credentials" > "Service Account"
   • Give it a name (e.g., "activepieces-vertexai")
   • Click "Create and Continue"
   • Grant the following role:
     • Vertex AI User - to use Vertex AI models
   • Click "Continue" and then "Done"

4. Create and Download Service Account Key

   • Click on the newly created service account
   • Go to the "Keys" tab
   • Click "Add Key" > "Create new key"
   • Choose JSON format
   • Click "Create" - this will download a JSON file

5. Add the Key to Activepieces

   • In Activepieces, paste the entire JSON content from the downloaded file into the "Service Account JSON Key" field
   • The piece will validate the credentials automatically

Actions

1. Generate Content (Gemini)

Generate responses using Google Vertex AI's Gemini model.

Parameters:

• Location (default: us-central1) - Google Cloud region where your Vertex AI resources are located
• Model (default: gemini-2.5-flash) - The Vertex AI model to use (e.g., gemini-pro, gemini-2.5-flash)
• System Message (optional) - Instructions to guide the model's behavior and tone
• User Message (required) - The prompt or question to send to the model

Example Output:

{
  "candidates": [
    {
      "content": {
        "role": "model",
        "parts": [
          {
            "text": "The model's response here..."
          }
        ]
      }
    }
  ]
}

2. Raw API Request

Make custom authenticated requests directly to the Vertex AI API for advanced use cases.

Parameters:

• Method (required) - HTTP method: GET, POST, PUT, PATCH, or DELETE
• URL (required) - Full Vertex AI API endpoint URL (e.g., https://aiplatform.googleapis.com/v1/projects/...)
• Body (optional) - JSON payload for the request

Example:

• Method: POST
• URL: https://aiplatform.googleapis.com/v1/projects/{projectId}/locations/us-central1/publishers/google/models/gemini-2.5-flash:generateContent
• Body: Custom JSON payload

Available Models

Common Gemini models available on Vertex AI:

• gemini-2.5-flash - Fast, cost-efficient model with thinking support
• gemini-2.5-pro - Most capable model for complex reasoning tasks
• gemini-2.0-flash - Balanced speed and capability
• gemini-1.5-flash - Fast and versatile
• gemini-1.5-pro - Best performing 1.5 generation model

For the full list of available models, check the Vertex AI documentation.

Common Use Cases

1. Generate Blog Content

Use the Generate Content action with a detailed prompt to automatically create blog posts based on a topic.

2. Customer Support Automation

Provide customer queries to Gemini for intelligent response suggestions.

3. Code Generation

Use the Generate Content action with code-related prompts to generate code snippets.

4. Data Analysis

Send data to Vertex AI for analysis and insights.

5. Custom API Integration

Use the Raw API Request action for advanced Vertex AI features not covered by the basic actions.

Error Handling

Common errors and solutions:

• Authentication Failed: Verify your Service Account JSON key is valid and has the necessary permissions
• API Not Enabled: Ensure Vertex AI API is enabled in your Google Cloud project
• Invalid Model: Check the model name matches available Vertex AI models in your region
• Location Not Found: Verify the specified location is available in your Google Cloud region

Best Practices

1. Use System Messages: Provide clear system messages to guide the model's behavior
2. Handle Rate Limits: Implement delays if making many requests in succession
3. Cache Responses: Store responses when possible to reduce API calls
4. Monitor Costs: Vertex AI API calls incur costs - monitor your usage in Google Cloud Console
5. Use Raw Request for Advanced Features: When specific features aren't available through the basic actions, use the Raw API Request action

Links

• Vertex AI Documentation
• Vertex AI API Reference
• Gemini Model Documentation
• Google Cloud Console