Usage & Costs in 📝MythOS is the reporting half of 📝Billing & Plans — what your library's AI chat actually cost this cycle, and where that cost came from. It reports spend that has already happened, so it answers where did the money go rather than what will this cost.
An owner-funded chat is only safe if the owner can see it. Because visitor conversations draw down your balance, an unexplained figure is the difference between confidently opening your library to the public and quietly turning chat off. Every number here breaks down until it names a model, a kind of visitor, or a single conversation.
It is written for creators running a public chat, and useful to anyone curious what their own AI use costs.
Key Capabilities
- Estimated usage this cycle — total spend for the current calendar month, with the platform fee stated
- Billing-cycle window — the exact date range the figures cover
- Budget tiles — your monthly, anonymous, and authenticated caps shown next to spend against them
- Cost over time — a daily spend trend across the last 30 days
- Cost by model — how the month's spend divides across the models actually used
- Cost by requester — spend split by who was talking: you, a contact, a signed-in visitor, or an anonymous one
- Thread attribution — cost broken down to the individual conversation
- Token counts — input and output tokens recorded alongside every cost figure
Getting Started
- Go to Settings → Billing, titled Billing & Usage
- Read This billing cycle for the headline figure and the dates it covers
- Scroll to Cost over time for the daily trend, then the Cost by model and Cost by requester panels for the breakdown
- Open Thread attribution to trace spend down to individual conversations
- To cap spend rather than just watch it, set your monthly limits under Budget Caps at Settings → AI
FAQ
- What period do these numbers cover? The headline figures cover the current calendar month and reset when it rolls over. The trend chart is a rolling 30 days and does not reset.
- Does the total include the platform fee? On MythOS-provided model keys, yes — the figure states the current fee percentage beside it. If you use your own provider key, the total shows raw provider cost and notes that your provider bills you directly.
- Can I see what each model charges per token? Not as a price list. This page reports what you spent on each model, not the underlying rate card.
- Why does a visitor's conversation appear in my usage? Chat spend is attributed to the library owner, so everything anyone spends talking to your library appears here.
- What happens when a budget cap is reached? Chat is blocked with a message explaining the cap, until the calendar month turns over or you raise the limit.
Related
- 📝Platform Credits — the balance these costs are drawn from
- 📝Chat Settings — configuring the caps and rate limits that bound this spend