Skip to main content
Prism Vertex offers two configuration modes to suit your deployment environment and authentication needs.

​
Configuration modes

You can configure Prism Vertex in one of two modes:

Standard mode

Full Vertex AI access with project ID, location, and service account authentication. Supports all model providers.

Express mode

Simplified API key-only configuration. Limited to Google Gemini models only.

​
How configuration works

Prism Vertex automatically detects which mode to use based on the configuration options you provide:
1

Check for project ID and location

If both project_id and location are present, the provider uses Standard mode.
2

Fall back to Express mode

If project_id or location are omitted, the provider automatically switches to Express mode endpoints.
3

Select authentication method

The provider authenticates using either your service account credentials or API key based on which option you configured.

​
Configuration location

Add your Vertex configuration to your Prism configuration file at config/prism.php: 
return [
    'providers' => [
        // Other providers...
    ],

    'vertex' => [
        // Your Vertex configuration here
    ],
];

​
Shared configuration

All Vertex AI model providers (Gemini, Anthropic, Meta, Mistral, etc.) read from the same vertex configuration block. You don’t need separate configurations for different model families.
If you need different settings for a specific provider (like a different region for Anthropic models), you can add a provider-specific override. See the Standard mode documentation for details.

​
Next steps

Standard mode setup

Configure with project ID, location, and service account credentials

Express mode setup

Configure with just an API key for quick setup

Authentication methods

Learn about service account and API key authentication

Build docs developers (and LLMs) love

Get started for freeTalk to us