Model techniques map
Techniquesinference & servingreasoning control

implementation detail · filed under inference & serving

Disabling reasoning via chat-template configuration

Reasoning or thinking is disabled through a chat-template or API parameter such as enable_thinking=false.

Also called disabling reasoning mode via chat template, Disable Reasoning, disabling reasoning/thinking mode, enable_thinking: False.

sources
5
models
2
lab adopt it
1
strongest
optional

How sources treat it

One count per evidence span, weakest treatment to strongest.

optional 5

Documented in

Evidence

5 spans quoted from the sources, strongest treatment first.

chat_template_kwargs": {"enable_thinking": False

optionalinference servingin Qwen3.5-122B-A10BQwen

You can obtain direct response from the model without thinking by configuring the API parameters.

optionalinference servingin Qwen3.5-35B-A3BQwen

Add --reasoning-parser qwen3 --default-chat-template-kwargs '{"enable_thinking": false}'to disable reasoning mode via command-line parameters.

optionalinference servingin vLLMvLLM

Disable reasoning: add --default-chat-template-kwargs '{"enable_thinking": false}'.

optionalinference servingin vLLMvLLM

add --default-chat-template-kwargs '{"enable_thinking": false}'

optionalunclearin Qwen3.6-35B-A3BQwen

Filed alongside

Other methods under inference & serving :: reasoning control.