Google Vertex AI Embedding Models

from: google:<model-id> selects a Vertex AI embedding model. A model ID, GCP project, location, and Google Cloud credentials are required. Google AI Studio API keys are not supported.

Parameter	Description	Default
google_project	Required. GCP project ID containing lowercase letters, digits, and hyphens.	-
google_location	Required. GCP region (for example us-central1) or global.	-
google_service_account_path	Path to a GCP service account JSON key file.	-
google_service_account_key	Service account JSON. Supports secret replacement.	-
google_application_default_credentials	Application Default Credentials from the file specified by GOOGLE_APPLICATION_CREDENTIALS.	false
dimensions	The output dimensionality of the embeddings. Some embedding models support dynamic output sizes.	-

Authentication requires exactly one of google_service_account_path, google_service_account_key, or google_application_default_credentials: true.

Below is an example configuration in spicepod.yaml:

embeddings:
  - from: google:gemini-embedding-001
    name: gemini_embeddings
    params:
      google_project: my-gcp-project
      google_location: us-central1
      google_service_account_path: /etc/spice/gcp-service-account.json
      dimensions: 768 # optional parameter

Vertex AI text embeddings lists model availability by project and region.

For detailed instructions and examples on running vector searches, refer to the Vector-Based Search documentation.