Skip to content
Discussion options

You must be logged in to vote

On current Jan builds you can make this a per-model default from the chat input:

  1. Select the model.
  2. Open the brain/Reasoning menu beside the prompt.
  3. Choose Off.

Jan persists that choice in the model's settings and, for llama.cpp models, sends this on each request:

{"chat_template_kwargs":{"enable_thinking":false}}

Auto omits the option and lets the model's chat template decide; On sends true.

If your older Jan version does not have that Reasoning menu, /no_think is only a per-prompt soft switch for Qwen-style templates. In that case update Jan, or set chat_template_kwargs.enable_thinking=false in the API request/model configuration if your backend exposes it.

The current default is auto…

Replies: 1 comment

Comment options

You must be logged in to vote
0 replies
Answer selected by cameronj86
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
2 participants