Allow specifying LLM version when initializing ChatGPT::Client

[andyklimczak]

Right now, when creating a new ChatGPT::Client, the model/LLM version being used seems fixed or implicit. For some use cases, it would be useful to explicitly specify which version of the model (e.g. gpt-4o-mini, gpt-4.1, gpt-3.5-turbo, etc.) the client should talk to.

Maybe something like

client = ChatGPT::Client.new(
  api_key: ENV["OPENAI_API_KEY"],
  model: "gpt-4o-mini"
)

The use case being I want to use specific models for different prompts for cost savings, etc.

Thanks, awesome gem!!

[nagstler]

@andyklimczak Thanks for raising the issue! Feel free to open a PR if you’re interested — I’m actively reviewing contributions.