with_thinking_mode

with_thinking_mode(
  self,
  enabled: bool = True,
  **kwargs: Any = {}
) ->

View source on GitHub

Name	Type	Description
`enabled`	`bool`	Default:`True`
`**kwargs`	`Any`	Default:`{}`

Example:

from langchain_nvidia_ai_endpoints import ChatNVIDIA

model = ChatNVIDIA(model="nvidia/nvidia-nemotron-nano-9b-v2")

# Enable thinking mode
thinking_model = model.with_thinking_mode(enabled=True)
response = thinking_model.invoke("Hello")

# Disable thinking mode
no_thinking_model = model.with_thinking_mode(enabled=False)
response = no_thinking_model.invoke("Hello")

LangChain Assistant

Menu

with_thinking_mode

Parameters