feat: add o3-mini model support and fix cost calculations

[cmann50]

Added support for o3-mini model including:

• Added o3-mini model to model_details.tmpl with correct token rates and specs:

  • Input: .10/1M tokens (/opt/homebrew/bin/bash.0000011 per token)
  • Output: .40/1M tokens (/opt/homebrew/bin/bash.0000044 per token)
  • Context length: 200k tokens
  • Support for function calling, vision, and parallel function calling

• Added temperature handling for o3/o1 models

  • Skip temperature parameter for o3/o1 models since they don't support it
  • Maintain temperature parameter for other models (e.g., gpt-4)

• Added support for reasoning_effort parameter

  • Added configuration in codai-config.yml
  • Added environment variable support (CHAT_REASONING_EFFORT)
  • Added command line flag support

• Code improvements:

  • Better cost calculation formatting
  • Cleaned up unused imports
  • Improved code organization

[meysamhadeli]

Added support for o3-mini model including:

• Added o3-mini model to model_details.tmpl with correct token rates and specs:

  • Input: .10/1M tokens (/opt/homebrew/bin/bash.0000011 per token)
  • Output: .40/1M tokens (/opt/homebrew/bin/bash.0000044 per token)
  • Context length: 200k tokens
  • Support for function calling, vision, and parallel function calling

• Added temperature handling for o3/o1 models

  • Skip temperature parameter for o3/o1 models since they don't support it
  • Maintain temperature parameter for other models (e.g., gpt-4)

• Added support for reasoning_effort parameter

  • Added configuration in codai-config.yml
  • Added environment variable support (CHAT_REASONING_EFFORT)
  • Added command line flag support

• Code improvements:

  • Better cost calculation formatting
  • Cleaned up unused imports
  • Improved code organization

Hi, Thanks for contribution :)
I had some comments, please take a look :)