STOP WASTING PAID TOKENS. START POOLING ACCOUNTS TODAY. [ GET YOUR VAULT ]

Set up Google Vertex AI

Who this is for: anyone who wants SpiderGate to use Gemini through their own Google Cloud (Vertex AI) project.

What this does: explains what SpiderGate's Vertex AI entry can and cannot do today, and shows the working alternative.

Before you begin

  • Read this first: SpiderGate's Google Vertex AI form asks only for a Project ID and a Location. It has no field for a Google Cloud credential.

  • Google's own docs say Vertex AI needs billing on the project and a credential, normally Application Default Credentials or a service account. A project ID alone does not let anyone call your project.

Limits and terms to know first

Warning: SpiderGate cannot sign in to your Google Cloud project for Vertex AI yet. A Vertex AI entry you add saves (SpiderGate runs no check for it) but cannot serve requests from your project. Until SpiderGate takes a credential, use a Google AI Studio key for Gemini.

Steps

Step 1 — Use Google AI Studio for Gemini

Follow Get a Google AI Studio key. It serves Gemini models through SpiderGate with a key from your own Google Cloud project.

Step 2 — Remove a Vertex AI entry you already added

Open Gate → Vault (/dashboard/gate/vault), find the Google Vertex AI key, and delete it, so no routing alias picks a slot that cannot answer.

Step 3 — Have your Project ID and Location ready for later

When SpiderGate supports a Vertex credential, you will need the Project ID (Google Cloud console Welcome page, after you pick the project) and a Location such as us-central1.

Verify it worked

Your vault no longer lists a Vertex AI key, and a Google AI Studio key appears instead.

Troubleshoot

Requests through a Vertex AI slot fail. That is expected today: SpiderGate holds no credential for your project. Use the Google AI Studio key from Step 1.

Related