On 31 August, Ollama introduced a new approach for its cloud service. Usage is measured in tokens – small pieces of text the model reads or writes – and linked to a monthly usage allowance.
If you are simply trying local AI, it helps to keep the two things separate. A model running on your own computer and one running in Ollama’s cloud use different resources. Cloud pricing does not describe the cost of every local conversation.
Ollama also says existing subscriptions remain in place unless the user chooses to switch. It is therefore worth reading what actually applies to your account.
The change points to something broader: as AI becomes an everyday tool, its use needs to be understandable beyond the chat itself. Clear limits and a good overview can make it easier to try a tool without guessing what the next step involves.