Skip to main content
Google Gemini offers powerful AI models with a generous free tier, making it an excellent choice for code analysis with Vibrant CLI without any upfront costs.

​
Why Gemini?

  • Free tier - Generous limits for developers at no cost
  • Fast - Quick response times for rapid analysis
  • High quality - Excellent code understanding capabilities
  • No credit card required - Start analyzing immediately

​
Get an API key

1

Visit Google AI Studio

Go to aistudio.google.com and sign in with your Google account.
2

Get API key

Click “Get API key” in the left sidebar.
3

Create key

Click “Create API key” and select a Google Cloud project (or create a new one).
4

Copy your key

Your key will look like AIza.... Copy it and keep it secure.
No credit card is required for the free tier. You can start using Gemini immediately.

​
Setup

Set the environment variable in your shell: 
export GOOGLE_GENERATIVE_AI_API_KEY="AIza..."
To make it permanent, add it to your shell profile: 
# Add this line to your shell profile
export GOOGLE_GENERATIVE_AI_API_KEY="AIza..."

​
Option 2: Alternative environment variable

You can also use the shorter GEMINI_API_KEY variable: 
export GEMINI_API_KEY="AIza..."

​
Option 3: .env file

Create a .env file in your project directory:
.env
GOOGLE_GENERATIVE_AI_API_KEY=AIza...
# Or alternatively:
# GEMINI_API_KEY=AIza...
Add .env to your .gitignore to prevent committing your API key.

​
Option 4: Configuration file

Create a vibrant.config.js file:
vibrant.config.js
module.exports = {
  provider: 'gemini',
  // API key will be read from environment variable
};

​
Usage

Run Vibrant with Gemini: 
# Auto-detect provider (if GOOGLE_GENERATIVE_AI_API_KEY is set)
vibrant . --ai

# Explicitly specify Gemini
vibrant . --ai --provider gemini

​
Available models

Vibrant uses gemini-2.0-flash-lite by default for optimal speed and quality.
  • Context: 1M tokens
  • Cost: Free tier available
  • Speed: Very fast
  • Quality: Excellent for code analysis
  • Best for: Daily development, CI/CD pipelines
  • Free tier: 15 RPM, 1M TPM, 1500 RPD
  • Context: 1M tokens
  • Cost: Free tier available
  • Speed: Very fast
  • Quality: High performance
  • Best for: Fast analysis of large codebases
  • Free tier: 15 RPM, 1M TPM, 1500 RPD
  • Context: 2M tokens
  • Cost: Free tier available
  • Speed: Fast
  • Quality: Superior reasoning
  • Best for: Complex analysis, large projects
  • Free tier: 2 RPM, 32K TPM, 50 RPD
RPM = Requests per minute, TPM = Tokens per minute, RPD = Requests per day

​
Change the model

Set the GEMINI_MODEL environment variable: 
export GEMINI_MODEL="gemini-1.5-pro"
vibrant . --ai --provider gemini
Or in your .env file:
.env
GOOGLE_GENERATIVE_AI_API_KEY=AIza...
GEMINI_MODEL=gemini-1.5-pro

​
Free tier limits

ModelRequests/minTokens/minRequests/day
gemini-2.0-flash-lite151M1500
gemini-1.5-flash151M1500
gemini-1.5-pro232K50
These limits are generally sufficient for:
  • Regular development workflows
  • Small to medium team usage
  • CI/CD pipeline integration
  • Learning and experimentation
If you need higher limits, you can enable billing in Google Cloud Console. Pay-as-you-go pricing is very affordable.

​
Example output

vibrant . --ai --provider gemini

🔮 Vibrant Analysis
─────────────────────

📡 AI Analysis (gemini:gemini-2.0-flash-lite)

📊 Token Optimization:
   Original: 38,000 tokens
   Summarized: 15,200 tokens
   Savings: 60.0%

✔ Analysis complete

📋 Summary:
The codebase has solid structure but contains security issues
that need immediate attention. Error handling could be improved
for better debugging and user experience.

🔍 Key Findings:
  • API credentials hardcoded in 2 files
  • 8 empty catch blocks that swallow errors
  • Debug console.log statements left in production code

💡 Recommendations:
  • Move all secrets to environment variables
  • Add proper error logging and monitoring
  • Remove debug code before deployment
  • Consider implementing a centralized logging service

✕ 2 errors · ⚠ 8 warnings · 75 files · 6s

​
Troubleshooting

​
Error: API key not found

Gemini API error: API key not valid
Solution: Set the GOOGLE_GENERATIVE_AI_API_KEY or GEMINI_API_KEY environment variable: 
export GOOGLE_GENERATIVE_AI_API_KEY="AIza..."

​
Error: Quota exceeded

Gemini API quota exceeded. Free tier limit reached.

Options:
  • Wait a few minutes and try again
  • Try Ollama (free, local): vibrant . --ai --provider ollama
  • Enable billing: https://aistudio.google.com/
Solutions:
  1. Wait and retry: Free tier resets per minute/day. Wait a few minutes.
  2. Use Ollama: Switch to local models: vibrant . --ai --provider ollama
  3. Use OpenRouter: Try free models: vibrant . --ai --provider openrouter
  4. Enable billing: Add payment method for higher limits (very affordable)

​
Error: Resource exhausted

Gemini API error: resource exhausted
Solution: You’ve hit the rate limit. Vibrant automatically retries, but if it persists: 
# Wait 60 seconds and try again
sleep 60 && vibrant . --ai --provider gemini

# Or use a different provider
vibrant . --ai --provider openrouter

​
Best practices

1

Use for daily development

Gemini’s free tier is perfect for frequent code analysis during development.
2

Respect rate limits

For large projects, analyze in smaller batches or use specific file patterns.
3

Enable billing for teams

If multiple team members are using the same API key, consider enabling pay-as-you-go billing.
4

Use in CI/CD

Gemini’s speed and free tier make it ideal for automated code quality checks.

​
CI/CD integration

For GitHub Actions:
.github/workflows/code-quality.yml
name: Code Quality

on: [push, pull_request]

jobs:
  analyze:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      
      - name: Setup Node
        uses: actions/setup-node@v4
        with:
          node-version: '20'
      
      - name: Run Vibrant with Gemini
        run: npx vibrant-cli . --ai --provider gemini
        env:
          GOOGLE_GENERATIVE_AI_API_KEY: ${{ secrets.GOOGLE_GENERATIVE_AI_API_KEY }}
Add GOOGLE_GENERATIVE_AI_API_KEY as a secret in your repository settings.
If you need higher limits, enable billing:
1

Enable billing

Visit Google Cloud Console and add a payment method.
2

Very affordable pricing

  • gemini-2.0-flash-lite: ~$0.01 per 1M tokens
  • gemini-1.5-flash: ~$0.04 per 1M tokens
  • gemini-1.5-pro: ~$1.25 per 1M tokens
3

Significantly higher limits

Paid tier removes most rate limiting for typical usage.

​
Next steps

AI providers overview

Compare all available providers

Ollama setup

Try local models for unlimited free usage

Build docs developers (and LLMs) love

Get started for freeTalk to us