Configuration Options

AI Models

CommitStudio leverages OpenAI's latest models to provide intelligent code analysis. Each model offers different capabilities, performance characteristics, and cost profiles:

The default model is gpt-4.1-mini, offering a good balance of performance and cost for most use cases. Here's how to select different models based on your needs:

Model Selection Guidelines

• For critical code reviews: Use gpt-4o or gpt-4.1 for the most thorough analysis
• For day-to-day development: Use gpt-4.1-mini for good results at reasonable cost
• For quick checks: Use gpt-4.1-nano when you need fast feedback
• For large repositories: Use models with higher token limits to analyze more code at once

Token Limits

Tokens are the basic units of text that the AI processes. Each model has different token limits that determine how much code can be analyzed at once:

# Set max tokens for responses
commitstudio config --max-tokens 3000
Copy code

Understanding Token Usage

• Code Analysis: More complex code requires more tokens to analyze
• Response Detail: Higher token limits produce more detailed analysis
• Cost Impact: Token usage directly affects API costs
• Context Window: Separate from max tokens, this is the model's capacity to "see" code

Recommended Token Limits

How Model Selection Affects Analysis

Different models analyze code differently:

• Higher-tier models (gpt-4o, gpt-4.1)

  • Recognize complex design patterns
  • Identify subtle bugs that might only appear in edge cases
  • Suggest architectural improvements
  • Provide detailed explanations of potential issues
  • Make more nuanced trade-off recommendations

• Mid-tier models (gpt-4.1-mini, o4-mini)

  • Catch common coding errors
  • Identify straightforward performance issues
  • Suggest code simplifications
  • Recommend basic best practices
  • Provide reasonable explanations

• Lower-tier models (gpt-4.1-nano, o3-mini)

  • Find syntax errors and obvious bugs
  • Suggest simple improvements
  • Provide brief explanations
  • Work best on smaller, simpler code segments

Configuration Commands

View Current Settings

To see your current configuration:

commitstudio config --view
Copy code

This displays:

• Current AI model
• Maximum tokens setting
• GitHub token status
• OpenAI API key status
• Cache configuration

Interactive Configuration

For a guided setup experience:

commitstudio config
Copy code

This launches an interactive prompt that:

1. Shows available AI models
2. Allows selection from a menu
3. Prompts for token limit settings
4. Validates your choices
5. Saves your configuration

Direct Configuration

For scripting or quick changes:

# Set a specific model
commitstudio config --model gpt-4o

# Set max tokens
commitstudio config --max-tokens 3000

# Configure multiple settings at once
commitstudio config --model gpt-4.1-mini --max-tokens 2500
Copy code

Configuration Persistence

Your configuration settings are stored securely on your machine:

• macOS: ~/Library/Preferences/commitstudio-nodejs
• Linux: ~/.config/commitstudio
• Windows: %APPDATA%\commitstudio-nodejs

These settings persist between sessions and can be reset with:

commitstudio --reset
Copy code

Related Topics

• Environment Variables - Configure CommitStudio using environment variables
• Customizing Behavior - Fine-tune how CommitStudio analyzes your code
• Standard Mode - See how configuration affects standard analysis mode