The Usage page at tesana.ai/settings/usage is the full audit trail for your credits. Every charge on your account appears here, per prompt, with enough detail to reconstruct exactly why it cost what it cost.

​
Summary cards

Across the top:
  • Credits used — total spend for the period.
  • Prompts — how many charges make up that spend.
  • Current streak and longest streak — consecutive days with at least one prompt. Purely for fun; streaks don’t affect pricing.

​
Model pricing reference

The same rate cards from the Pricing page, inline — so any charge in the ledger below can be sanity-checked against the model’s published rates without leaving the page.

​
Days building on Tesana

A GitHub-style activity grid of the past year. A day lights up when you sent at least one prompt.

​
By model / By game

Two rollup tables:
  • By model — credits, prompt count, and cache hit % per model. The cache hit rate is worth watching: high (90%+) means your prompts are re-reading project context at the cheap cached rate, which is the healthy state.
  • By game — credits and prompts per game, so you can see which project is consuming your balance.

​
The charges ledger

The heart of the page: one row per charge, newest first. Each row shows the prompt text, the game, the model, the date, and the credits charged. A negative figure is a refund — it appears when a provider reports a lower cost than we first estimated. Expand any row for the receipt-level detail:
To sanity-check any charge, take the expanded token counts and multiply by the per-model rates.

​
A note on partial charges

A single long session can appear as more than one ledger row — for example a partial charge when the AI pauses to ask you a question, then a final charge when the build completes. The rows always sum to the session’s actual total; nothing is double-billed.