Google Gemini Setup Guide

Quick guide for using Google Gemini with doctai.

Getting Your API Key

  1. Go to Google AI Studio
  2. Sign in with your Google account
  3. Click “Create API Key”
  4. Copy your API key

Configuration

Method 1: Environment Variable 

export DOCTAI_API_KEY="your-gemini-api-key-here"
export DOCTAI_PROVIDER="gemini"

Method 2: Command Line 

doctai --docs README.md \
  --api-key "your-gemini-api-key" \
  --provider gemini

Usage Examples

Basic Usage 

doctai --docs README.md \
  --api-key $DOCTAI_API_KEY \
  --provider gemini

With Specific Model 

doctai --docs README.md \
  --api-key $DOCTAI_API_KEY \
  --provider gemini \
  --model gemini-1.5-pro-latest

Multiple Documents 

doctai --docs README.md INSTALL.md \
  --api-key $DOCTAI_API_KEY \
  --provider gemini

With Output File 

doctai --docs README.md \
  --api-key $DOCTAI_API_KEY \
  --provider gemini \
  --output results.json

Available Models

Gemini provides several models. The default is gemini-1.5-pro-latest, but you can also use:

  • gemini-1.5-pro-latest - Latest Gemini 1.5 Pro (default)
  • gemini-1.5-flash-latest - Faster, lighter model
  • gemini-pro - Previous generation

Example with Flash model:

doctai --docs README.md \
  --api-key $DOCTAI_API_KEY \
  --provider gemini \
  --model gemini-1.5-flash-latest

GitHub Actions Integration

To use Gemini in GitHub Actions:

  1. Add your Gemini API key as a repository secret:
    • Go to Settings → Secrets and variables → Actions
    • Click “New repository secret”
    • Name: GEMINI_API_KEY
    • Value: Your Gemini API key
  2. Use the workflow with Gemini provider:
name: Test Documentation

on: [push]

jobs:
  test-docs:
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@v4
    - uses: actions/setup-python@v5
      with:
        python-version: '3.11'
    - run: pip install doctai
    - run: |
        doctai \
          --docs README.md \
          --api-key $ \
          --provider gemini

Or use the included workflow with manual trigger:

  1. Go to Actions → Test Documentation
  2. Click “Run workflow”
  3. Select gemini as the provider
  4. Click “Run workflow”

Example Script

Use the provided example script:

export GEMINI_API_KEY="your-api-key"
./examples/gemini-example.sh

API Costs

Google Gemini offers competitive pricing:

  • Gemini 1.5 Pro:
    • Free tier: 2 requests per minute
    • Paid: ~$0.001-0.007 per 1K characters
  • Gemini 1.5 Flash:
    • Free tier: 15 requests per minute
    • Lower cost than Pro

Check the latest pricing at: https://ai.google.dev/pricing

Troubleshooting

“Failed to communicate with Gemini”

Solution: Check that:

  • Your API key is correct
  • You have API access enabled
  • You haven’t exceeded rate limits

“Unexpected response format from Gemini API”

Solution:

  • Verify you’re using a supported model
  • Check your API key has proper permissions
  • Try with the default model (gemini-1.5-pro-latest)

Rate Limits

Free tier has rate limits:

  • 2 requests/minute for Pro
  • 15 requests/minute for Flash

Solution:

  • Wait and retry
  • Upgrade to paid tier
  • Use Flash model for testing

Comparison with Other Providers

Feature OpenAI Anthropic Gemini
Cost $$$ $$$ $
Speed Fast Fast Very Fast (Flash)
Context 128K 200K 1M+
Free Tier Limited Limited Yes

Gemini is great for:

  • ✅ Cost-effective testing
  • ✅ Very large documentation
  • ✅ Fast iteration with Flash model
  • ✅ Experimentation with free tier

Best Practices

  1. Start with Free Tier: Test with the free tier before upgrading
  2. Use Flash for Development: Faster and cheaper for testing
  3. Use Pro for Production: More capable for complex documentation
  4. Monitor Usage: Keep track of API usage to avoid surprises
  5. Cache Results: Use --output to save results and avoid re-runs

Additional Resources

Happy Testing with Gemini! 🚀