Skip to content
Documentation
Getting Started
Authenticate

Authenticate

Cogo speaks Gemini through one of two auth paths. Pick whichever matches the credentials you have.

Option A — Public Gemini API

Best for personal use and quick experiments.

  1. Get an API key from Google AI Studio.
  2. Export it:
export GOOGLE_API_KEY=...

That’s it. Cogo will detect the key on startup and use the public Gemini endpoints.

Option B — Vertex AI

Best for org-managed projects, longer context windows, and tighter audit trails. Requires a GCP project with the Vertex AI API enabled.

  1. Authenticate with gcloud:
gcloud auth application-default login
  1. Tell Cogo to use Vertex:
export GOOGLE_GENAI_USE_VERTEXAI=true
export GOOGLE_CLOUD_PROJECT=your-project
export GOOGLE_CLOUD_LOCATION=us-central1   # or "global"

The same Application Default Credentials (ADC) flow that other GCP tooling uses.

Verify 

cogo -p "ping"
# pong

If you see an error, double-check that exactly one auth path is configured (e.g. don’t have both GOOGLE_API_KEY and GOOGLE_GENAI_USE_VERTEXAI set).

A copy-pasteable template

The repo ships an .env.example you can drop in as a starting point — copy it to .env, edit the values, and source .env before running.

Next

→ First Prompt — run a turn.

InstallFirst Prompt