Closed vlad-isan closed 4 months ago
Have you tried using the transient menu to set the backend/model? (C-u M-x gptel-send
)
It works if I use the transient menu, was just wondering if it was possible to preset the backend and model on a keybinding, and open gptel with a specific backend & model, without using the transient menu.
This would serve as a shortcut to have predefined keybindings for different backends / models, one for gpt4-turbo, one for ppx online, without past messages / context for example.
It works if I use the transient menu, was just wondering if it was possible to preset the backend and model on a keybinding, and open gptel with a specific backend & model, without using the transient menu.
When you set the model/backend in the transient menu, it's persistent in that buffer. Do you often need to switch backends within a single buffer?
This would serve as a shortcut to have predefined keybindings for different backends / models, one for gpt4-turbo, one for ppx online, without past messages / context for example.
This is doable with some elisp. See #189.
That was helpful, thank you. Thanks for the quick replies
Hi there and many thanks for this amazing package!
Is there a way to set bindings to open gptel with different specific backends / models selected, without setting the backend as a global default?
It works when I set the backend / model with
setq-default
, but I couldn't make it work per invocation.Is there a way?