Skip to main content

Usage and billing

Settings → Organization → Usage & Cost shows how much your organisation is using FYAI and what it costs. It has three tabs:

TabShows
BillingYour token quota, plan, billing cycle, pay-as-you-go balance, available plans, daily usage, overage mode, balance transactions and usage by model.
By agent & projectToken usage broken down by agent and by project.
AlertsUsage alert rules.

Everyone in the organisation can open Usage & Cost. The figures cover the whole organisation, not just your current project.

To change plan, top up or cancel, see Plans.

How usage is measured​

FYAI records a usage entry each time a model is called for a chat turn, an agent run or a workflow step. Each entry records the input and output tokens, the model, the agent, the project and, for chat, the person.

Cost is calculated from per-model token prices (per million input and output tokens) kept by FYAI. A model call that FYAI can't price is recorded at zero cost.

There are two kinds of model:

  • Platform-managed models run on FYAI's provider keys. Their tokens count against your plan's quota and, past the quota, your pay-as-you-go balance. Your plan's markup applies to their cost.
  • Bring-your-own-key (BYOK) models use your own provider credentials. You pay the provider directly. BYOK tokens don't count against your quota or balance.
note

The Token Usage card counts platform-model tokens only. The daily chart, Usage by Model and By agent & project count all model calls, including BYOK models, so their totals can be higher. Est. Cost for BYOK calls is an estimate of what the provider charges you, not something FYAI bills.

Billing tab​

The page is headed Usage & Billing. Select Refresh to reload it.

Token Usage, Current Plan and Billing Cycle​

CardShows
Token UsageTokens used of your quota this cycle, with a progress bar that turns amber at 70% and red at 90%, and tokens remaining.
Current PlanYour plan, whether usage past the quota is allowed (Overage allowed) or blocked (Hard limit), and the overage rate per million tokens.
Billing CycleThe date your quota resets, and the estimated overage so far.

If your quota runs out, a banner explains what's happening: either usage is now drawing on your pay-as-you-go balance, or requests are blocked until the quota resets or you top up or upgrade. While usage is blocked, chats, agent runs and workflows that use platform-managed models stop with an error.

If a card payment fails, a banner shows how long your plan stays active and offers Update payment method. If it isn't fixed by then, the organisation moves to the Free plan. Your data is kept.

Pay-As-You-Go Balance​

A prepaid balance in dollars that pays for platform-model usage:

  • On Starter, every platform-model call draws from the balance. When it reaches zero, platform-model usage pauses until you top up.
  • On plans with a monthly quota, the balance is used once the quota is spent, if your overage mode is Switch to pay-as-you-go.

To add credit, select a preset Top up amount, or enter a Custom amount within the limits shown and select Top up. You pay through Stripe Checkout.

Available Plans​

A table of plans with Token Quota, Price and Billing Cycle, your current plan marked Current, and buttons to Upgrade, Switch, Cancel Subscription or Keep Subscription. Compare all plans opens the plans page. See Plans.

Daily Token Usage (Last 30 Days)​

A chart of total tokens per day, with days in UTC.

Overage Billing Mode​

Shown on paid plans. Choose what happens when your monthly quota is used up:

OptionBehaviour
Switch to pay-as-you-goRequests keep working past the quota and are paid for per token from your pay-as-you-go balance.
Hard stopRequests are blocked once the quota is used up. No extra charges.

Starter has no included quota, so it always works from the balance and this setting has no effect.

Balance Transactions​

The last 50 changes to your balance: Date, Type (Top-up, Usage or Adjustment), Detail, Tokens, Amount and the resulting Balance. Usage covered by your quota appears as zero-amount Usage rows.

Usage by Model (Last 30 Days)​

Totals for Total Calls, Total Tokens, Input Tokens and Est. Cost, then a table per model with calls, input, output and total tokens and estimated cost, most expensive first.

Manage billing​

On paid plans, Manage billing opens Stripe's customer portal, where you manage your payment details and billing history. You return to Usage & Cost afterwards.

By agent & project tab​

Headed Billing Analytics, for the last 30 days:

  • Token Usage by Agent — a bar chart and table of Agent, Tokens and Calls.
  • Token Usage by Project — a donut chart and table of Project, Tokens, Est. Cost and Calls.

Each shows up to 10 rows. Projects are identified by project code, and agents by ID where a name isn't available. "Calls" counts recorded usage entries, which can be fewer than individual model requests.

Alerts tab​

Usage Alerts lets you define thresholds as a percentage of your token quota.

  1. Select New Alert (or Create first alert).
  2. Set Threshold (%), from 1 to 100 (80 by default).
  3. Choose a Channel: Email or Webhook. For a webhook, enter the Webhook URL.
  4. Select Create.

The table lists each alert's threshold, channel, status (Active or Disabled) and Last Notified time. Use the row actions to disable, enable, edit or delete an alert. Any member can manage alerts.

caution

Alert rules are saved, but FYAI does not yet send notifications for them; Last Notified stays at "Never". Until notifications are available, watch the Token Usage card or the usage meter in the header. Percentage alerts also don't apply on Starter, which has no included quota.

Usage meter​

On the Free plan, a meter in the header shows the percentage of your free tokens used. It turns amber at 70%. Select it to open Usage & Cost.