I have a quick question: how can I tell when Claude 4 Opus is being billed on a usage-based model and when it’s included in the Pro plan? I’m still a bit confused about how this works. I tried it for the first time today — up until now, I’ve only used Claude Sonnet.
I’m hopeful that we can add an in-product indicator for this. We raised it as a feature request with the engineering team last week. Definitely agree there should be more clarity in-product for this. I believe there currently isn’t an indicator for when usage-based pricing will be consumed in-product.
As a workaround, the usage tab should be updated live, so you can refresh the page to double check when you’re dipping into usage-based pricing. Generally with MAX and the most expensive models, you’ll hit this much more quickly. I’d recommend avoiding max and passing a bunch of tokens that you don’t necessarily need to, in order to cut down on token usage and your compute usage (to avoid hitting rate limits)