Using MoltBot with Google Gemini

Google Gemini offers powerful AI capabilities with competitive pricing. This guide covers setting up MoltBot with Gemini models.

Why Gemini?

Generous Free Tier: Free usage for Gemini 1.5 Flash
Long Context: Up to 1M tokens context window
Fast Response: Gemini Flash is optimized for speed
Multimodal: Supports text, images, and video
Cost Effective: Competitive pricing for Pro models

Getting Your API Key

Step 1: Access Google AI Studio

Go to Google AI Studio
Sign in with your Google account
Click "Get API Key"
Create a new API key or use existing one

Step 2: Save Your Key

# Set environment variable
export GOOGLE_AI_API_KEY="your-api-key-here"

Or add to your shell profile:

echo 'export GOOGLE_AI_API_KEY="your-api-key-here"' >> ~/.zshrc
source ~/.zshrc

Configuration

Basic Setup

Configure MoltBot to use Gemini:

moltbot config set aiProvider google
moltbot config set googleModel gemini-1.5-flash

Or edit ~/.moltbot/config.json:

{
  "aiProvider": "google",
  "google": {
    "apiKey": "your-api-key-here",
    "model": "gemini-1.5-flash",
    "maxTokens": 8192
  }
}

Available Models

Model	Best For	Context	Speed
`gemini-1.5-flash`	Fast responses, cost-effective	1M tokens	Fast
`gemini-1.5-pro`	Complex reasoning, analysis	1M tokens	Medium
`gemini-1.0-pro`	Legacy support	32K tokens	Medium

Model Selection

Choose based on your needs:

# For fast, everyday tasks
moltbot config set googleModel gemini-1.5-flash

# For complex analysis
moltbot config set googleModel gemini-1.5-pro

Advanced Configuration

Temperature and Creativity

{
  "google": {
    "model": "gemini-1.5-flash",
    "temperature": 0.7,
    "topP": 0.9,
    "topK": 40
  }
}

Safety Settings

Configure content filtering:

{
  "google": {
    "safetySettings": {
      "harassmentThreshold": "BLOCK_ONLY_HIGH",
      "dangerousContentThreshold": "BLOCK_ONLY_HIGH"
    }
  }
}

Options:

BLOCK_NONE
BLOCK_ONLY_HIGH
BLOCK_MEDIUM_AND_ABOVE
BLOCK_LOW_AND_ABOVE

Long Context

Leverage Gemini's 1M token context:

{
  "google": {
    "model": "gemini-1.5-flash",
    "maxInputTokens": 1000000
  },
  "memory": {
    "maxContextTokens": 500000
  }
}

Multimodal Capabilities

Image Understanding

Gemini can analyze images:

moltbot chat "What's in this image?" --image /path/to/image.jpg

Document Analysis

Process long documents:

moltbot analyze /path/to/document.pdf

Usage Examples

Chat with Gemini

# Start interactive chat
moltbot chat

# One-off query
moltbot ask "Explain quantum computing in simple terms"

With Messaging Platforms

Gemini works with all MoltBot integrations:

# Telegram with Gemini
moltbot config set telegram.enabled true
moltbot config set aiProvider google

# Start
moltbot start

Cost Optimization

Free Tier Usage

Gemini 1.5 Flash has a generous free tier:

15 requests per minute
1 million tokens per minute
1,500 requests per day

Monitoring Usage

Check your usage:

moltbot stats usage

Cost Saving Tips

Use Flash for simple tasks: Default to gemini-1.5-flash
Cache responses: Enable caching for repeated queries
Compress context: Summarize long conversations

{
  "google": {
    "model": "gemini-1.5-flash"
  },
  "cache": {
    "enabled": true,
    "ttl": 3600
  },
  "memory": {
    "summarizeAfterTokens": 50000
  }
}

Comparison with Other Providers

Feature	Gemini	Claude	GPT-4
Free Tier	Yes	No	No
Max Context	1M	200K	128K
Speed	Fast	Medium	Medium
Image Support	Yes	Yes	Yes
Code Quality	Good	Excellent	Excellent

Switching Between Models

MoltBot supports multiple AI providers. You can switch dynamically:

# Use Gemini for fast tasks
moltbot ask "Quick question" --provider google --model gemini-1.5-flash

# Use Claude for complex tasks
moltbot ask "Complex analysis" --provider anthropic --model claude-3-sonnet

Or set per-skill:

{
  "skills": {
    "quickAnswer": {
      "aiProvider": "google",
      "model": "gemini-1.5-flash"
    },
    "deepAnalysis": {
      "aiProvider": "anthropic",
      "model": "claude-3-opus"
    }
  }
}

Troubleshooting

API Key Errors

Error: Invalid API key

Solution:

Verify key at Google AI Studio
Check for extra spaces or newlines
Regenerate key if needed

Rate Limit Errors

Error: Rate limit exceeded