Skip to main content
Standard mode gives you full access to all Vertex AI model providers and features. It requires a Google Cloud project ID, location, and authentication credentials.

​
Configuration options

Add the following to your Prism configuration at config/prism.php: 
'vertex' => [
    'project_id'  => env('VERTEX_PROJECT_ID'),
    'location'    => env('VERTEX_LOCATION', 'us-central1'),
    'credentials' => env('VERTEX_CREDENTIALS'), // path to service-account.json
],
You can use either credentials (service account file path) or api_key for authentication. Service accounts are recommended for production environments.

​
Required parameters

​
project_id

Your Google Cloud project ID. This is the unique identifier for your GCP project, not the project name or number. 
'project_id' => env('VERTEX_PROJECT_ID'),
Set the environment variable: 
VERTEX_PROJECT_ID=your-project-id

​
location

The Google Cloud region where your Vertex AI resources are provisioned. Common options include:
  • us-central1 (Iowa)
  • us-east4 (Northern Virginia)
  • europe-west1 (Belgium)
  • asia-northeast1 (Tokyo)
'location' => env('VERTEX_LOCATION', 'us-central1'),
Set the environment variable: 
VERTEX_LOCATION=us-central1
Different regions may have different model availability and pricing. Check the Vertex AI locations documentation for details.

​
Authentication options

​
Complete configuration example

Here’s a complete Standard mode configuration: 
'vertex' => [
    'project_id'  => env('VERTEX_PROJECT_ID'),
    'location'    => env('VERTEX_LOCATION', 'us-central1'),
    'credentials' => env('VERTEX_CREDENTIALS'),
],
With the corresponding .env file: 
VERTEX_PROJECT_ID=my-gcp-project
VERTEX_LOCATION=us-central1
VERTEX_CREDENTIALS=/var/secrets/vertex-service-account.json

​
Per-provider overrides

All Vertex AI providers share the same vertex configuration by default. If you need different settings for a specific provider (such as using a different region for Anthropic models), add a provider-specific override: 
'vertex' => [
    'project_id'  => env('VERTEX_PROJECT_ID'),
    'location'    => env('VERTEX_LOCATION', 'us-central1'),
    'credentials' => env('VERTEX_CREDENTIALS'),
],

// Override for Anthropic only
'vertex-anthropic' => [
    'location' => 'europe-west1',
],
The provider-specific configuration inherits all options from the main vertex config and overrides only the specified values.

​
Available provider keys

  • vertex-gemini
  • vertex-anthropic
  • vertex-meta
  • vertex-mistral
  • vertex-deepseek
  • vertex-ai21
  • vertex-kimi
  • vertex-minimax
  • vertex-openai
  • vertex-qwen
  • vertex-zai

​
Supported model providers

Standard mode supports all Vertex AI model providers:
  • Google Gemini - Native Google models
  • Anthropic Claude - Claude 3 and 3.5 models
  • Meta Llama - Llama 3 and 4 models
  • Mistral - Mistral and Codestral models
  • DeepSeek - DeepSeek V3 models
  • AI21 - Jamba models
  • Kimi - Kimi K2 models
  • MiniMax - MiniMax models
  • OpenAI OSS - Open source GPT models
  • Qwen - Qwen 2.5 models
  • ZAI - GLM models
Express mode only supports Google Gemini models. Use Standard mode if you need access to partner model providers.

​
Next steps

Authentication

Learn more about service account and API key authentication

Express mode

See how to configure API-key-only Express mode

Build docs developers (and LLMs) love

Get started for freeTalk to us