Skip to main content

Vertex AI

Google Cloud Vertex AI serves Gemini and other models. Unstract authenticates to it with a Google Cloud service account key.

Getting started with Vertex AI​

  1. In your Google Cloud project, enable the Vertex AI API.
  2. Create a service account and download its JSON key. Refer to the GCloud docs on creating service account credentials.
  3. Note your project ID and the model name you want to use.

Setting up Vertex AI in Unstract​

  • Sign in to the Unstract Platform.
  • From the side navigation menu, choose Settings 🞂 LLMs.
  • Click on the New LLM Profile button.
  • From the list of LLMs, choose Vertex AI.
  • Name — a unique name for this adapter instance.
  • Credentials — paste the contents of your service account JSON key.
  • Project — your Google Cloud project ID.
  • Model — the Vertex AI model name (for example gemini-2.0-flash).
  • Location — the Google Cloud region for the endpoint (for example us-central1, global). Defaults to us-central1 if left empty.
  • Safety Settings — optional per-category harm filters; each defaults to BLOCK_ONLY_HIGH.
  • Enable Extended Thinking — for supported Gemini models; reveals a Thinking Budget Tokens field (use -1 for a dynamic budget).
  • Leave Max Retries and Max output tokens at their defaults.
  • Click on Test Connection and ensure it succeeds, then click Submit.