Hi Victor,

I understand the frustration. v0's docs say credits work like a prepaid balance: generations draw from your credit balance, and when credits are used up, generation pauses.

The part that may explain the fast usage is that v0 pricing is token-based. The docs say token usage varies based on the length of your input and the size of the output, and that v0 can include relevant context such as chat history, source files, and Vercel-specific knowledge as input. So a session that looks like "basic prompting" can still become expensive if the chat/project context is large or the generated output is substantial.

The best thing to check is the v0 Usage page, because the docs say it provides a detailed log of each event with the date, user, event type, model, and cost. That should show exactly where the $5 went and whether it was expected usage or something unusual.

I do agree that clearer in-product cost visibility would help. Even if the billing is technically correct, it would be useful to see an estimated cost before generation or a clearer breakdown after each generation, especially for hobby users on the free credits.

If you can share a screenshot of the Usage page with private details hidden, Vercel staff should be able to tell whether this was normal token usage, a specific model cost, or something abnormal.