You can check the pricing of models you are using in the respective API documentation. Pricing is usually based on 1M input/output tokens.
A future update will include a feature to see token usage in the app, but for now you can monitor usage in the respective API dashboards.