Gemini API Key Setup (Journal)

This flow is for Gemini API keys in Google AI Studio.

1. Prerequisites

A signed-in Google account.
Your account meets AI Studio access requirements (supported region and age requirements).
If your region is unsupported, use Gemini through Vertex AI.

2. Create the API key

Open AI Studio API Keys.
Select an existing project, or create one and generate a key.
Save the key securely. Use it only on server side and never commit it to code.

Note: new users are often given a default project and API key automatically.

3. Enable paid tier (if needed)

Review your current tier in Gemini API Billing docs.
In AI Studio API keys or project settings, click Set up billing for the target project.
After Cloud Billing is connected, that project’s keys use paid-tier billing and quotas.

4. Local environment variable

export GEMINI_API_KEY="your_api_key_here"

The Gemini SDK supports both GEMINI_API_KEY and GOOGLE_API_KEY. If both are set, GOOGLE_API_KEY takes precedence.

5. Minimal verification

curl "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.5-flash:generateContent?key=${GEMINI_API_KEY}" \
  -H "Content-Type: application/json" \
  -X POST \
  -d '{
    "contents": [{
      "parts": [{
        "text": "Return OK"
      }]
    }]
  }'

If the response contains model output, the key is working.

6. Common issues (direct cause -> direct fix)

Symptom: AI Studio redirects to region availability page
- Cause: current account region or eligibility requirements are not met.
- Fix: use an eligible account/region, or switch to Vertex AI.
Symptom: rate limits are hit quickly even with multiple keys
- Cause: limits are enforced at project level, not per key.
- Fix: review project-level quotas and upgrade tier or split projects if required.
Symptom: Free Tier works for testing but fails under production load
- Cause: Free Tier has lower quota and stricter limits.
- Fix: move to Paid Tier and monitor project quota usage.