DeepSeek V4.1 Flash now has effort levels
DeepSeek V4.1 Flash now has effort levels. In the model selector, the Effort row opens Low, High and Max, with High marked as recommended: it is the one used unless you pick another, so the model now thinks before it answers on every request.
Admins can cap it in Organization → Models with Maximum effort, like any other model. More about effort in Writing a request and about your organization’s models in Which models your organization uses.