Agents and models
Every cloud task runs one agent with one of its models. Every agent runs on the JetBrains AI provider, and your tasks consume AI credits based on the agent and model you select.
Which agents and models you can pick is decided for your organization in JetBrains Console: the agents come from your AI policy, and the models from the organization's list of JetBrains AI models. If an agent is missing, the selector shows Some agents are restricted by your organization. See Supported agents for the full lineup.
Select an agent and model for a new task
On the New Task page, click the agent selector under the task input (for example, Claude) and select the agent.
Click the model selector next to it (for example, Opus 5 Medium) and select a model of that agent.
The model you choose affects cost. More capable models usually use more tokens, which consumes more of your AI credits.
If the agent supports it, set Effort at the bottom of the same menu.
Effort controls how much reasoning the model uses for the task. Higher effort levels can produce more carefully reasoned results, but they usually take longer and use more tokens.
You can switch the agent only when you create a new task. Once the task is started, you can change the model, but not the agent. This is useful if you want a faster model for iteration or a stronger model for complex changes.
Change the model for an existing task
Open the task.
Under the task input, click the model selector.
Select another model of the same agent.