The description of get_max_tokens on this page is "This returns the maximum number of tokens allowed for the given model.".
The documentation could easily be (mis)interpreted as meaning the model's context window, but in reality the function returns max_tokens, which is basically an alias of max_output_tokens.
This description could be clarified to remove the ambiguity, e.g. by changing the sentence to "This returns the maximum number of output tokens allowed for the given model."
The description of
get_max_tokenson this page is "This returns the maximum number of tokens allowed for the given model.".The documentation could easily be (mis)interpreted as meaning the model's context window, but in reality the function returns
max_tokens, which is basically an alias ofmax_output_tokens.This description could be clarified to remove the ambiguity, e.g. by changing the sentence to "This returns the maximum number of output tokens allowed for the given model."