Currently, the pre-built Code Assistant is limited to using Claude Sonnet 3.5 as its default language model. While Claude Sonnet 3.5 is powerful, it can be overkill and expensive for simpler or more trivial coding tasks. It would be cool to be able to select a cheaper model for such tasks.
There are benefits such as:
- Reduce Costs: Use cheaper models for less complex tasks, saving on API expenses.
- Customize Workflows: Tailor the assistant to specific needs by choosing models optimized for certain tasks or languages.
- Experiment with Alternatives: Test and compare different LLMs to find the best fit for their use case.
Maybe an additional dropdown menu can be shown if the Code assistant is selected to be able to choose the model?
