Problem
Currently, the maximum supported tokens parameter is being input by the user via environment variables. This is not only an added overhead when the user wants to user different models, but also not needed.
Solution
Pydantic AI provided max token limits directly. This implementation can be swapped out with their function.
Problem
Currently, the maximum supported tokens parameter is being input by the user via environment variables. This is not only an added overhead when the user wants to user different models, but also not needed.
Solution
Pydantic AI provided max token limits directly. This implementation can be swapped out with their function.