Integrations
Google Vertex AI
How to get your Google Vertex AI API key and connect it to Nova Admin.
Connect a GCP project for Vertex AI by entering its projectId and a region (location). Authentication uses Application Default Credentials — enable the Vertex AI API and grant the runtime service account Vertex AI User on the project beforehand.
Prerequisites
- A Google Vertex AI account. Sign up at https://console.cloud.google.com/vertex-ai?project=nova-prod.
- An active Nova Admin workspace. Go to /agents/integrations to connect.
Steps
- Open the GCP console for the project that will host Vertex AI calls.
- Enable the
aiplatform.googleapis.com(Vertex AI) API on that project. - Grant the Cloud Functions runtime service account (e.g.
162939303532-compute@developer.gserviceaccount.com) theVertex AI Userrole. - Enter the projectId here (e.g.
nova-prod). - Set a location/region (e.g.
us-central1) where the models you use are available.
Credential fields
| Field | Required | Description |
|---|---|---|
| GCP Project ID | Yes | string (e.g. e.g. nova-prod) |
| Location (region) | No | string (e.g. e.g. us-central1) |
| Service Account Key JSON (optional) | No | Secret — stored encrypted, never exposed. (e.g. Paste the SA key file JSON. Only needed where the runtime has no ambient credentials (e.g. Netlify SSR); leave empty to use Application Default Credentials.) |
Troubleshooting
- Invalid credentials: verify the key was copied in full with no trailing spaces.
- Key not working: check that your Google Vertex AI account has API access enabled and billing is active if required.
- Need help?: refer to the Google Vertex AI documentation.