joeyeamigh I went to create a bang to activate Claude 3.7 Sonnet in Assistant, and noticed that the bang will always use extended thinking rather than the base model. I would love a way to differentiate the two. I would think that there would be a different "profile" parameter for 3.7 and 3.7 Thinking. If there is, then I would love to see it documented on https://help.kagi.com/kagi/ai/assistant.html#url-parameters.
__ In another post someone commented a website that showed some Kagi bangs. I don't know why it's on this website: https://kbe.smaertness.net/search?q=%2Fassistant I found that ?profile=claude-3-sonnet&q=%s will use 3.7 (not thinking)
Rehan Sorry about that, the docs were incorrect. claude-3-sonnet points to 3.7 without thinking, claude-3-7-sonnet points to 3.7 sonnet with extended thinking. We'll correct the docs, thanks for reporting!
