---
title: 'Usage'
description: 'Read project and team usage: included allocations, on-demand charges, agent activity and AI model spend.'
---

# Usage

Every project has a **Usage** page, and every team has one covering all of its projects. Use them to see how close a project is to its [included usage](/docs/pricing#included-usage), what it has cost so far this month, and where AI credit goes.



## Choose the period

Pick a date range at the top. **Current Billing Cycle** covers the calendar month so far (UTC), the period that included usage and on-demand charges are counted in. Use it when a project is close to a limit or has been restricted.

## Sections

| Section               | What it shows                                                                                                                                                                                                      |
| --------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
| **Overview**          | A table of every resource (API requests, capacity, storage, chat turns, automation runs, agent steps, sandbox compute, members, webhooks) with its **Included** amount, **On-demand** usage and **Charge** so far. |
| **Spend by Model**    | AI credit spend per AI model.                                                                                                                                                                                      |
| **Requests by Model** | Model calls per AI model.                                                                                                                                                                                          |
| **All Tokens**        | Input and output tokens over time.                                                                                                                                                                                 |
| **Agent Activity**    | Chat turns, automation runs and agent steps over time.                                                                                                                                                             |
| **API Requests**      | Requests to the project's APIs over time.                                                                                                                                                                          |
| **DB Capacity**       | Database size over time.                                                                                                                                                                                           |
| **Assets Storage**    | Asset storage over time.                                                                                                                                                                                           |

The team **Usage** page shows the same sections, adding up every project in the team.

## Read the numbers

Each chart uses its own unit: requests, tokens, gigabytes or dollars. Compare a chart only with its own allocation.

- **AI model spend** (Spend by Model, Requests by Model, All Tokens) is paid from [AI credits](/docs/billing/ai-credits). It doesn't count toward the plan's included usage.
- **Database capacity and asset storage** show size at each point in time. On Pro, the amount above the allocation is billed as a [time-weighted average](/docs/pricing#how-usage-is-calculated), not at its peak.
- **Agent steps** count each model call the agent makes. A long task uses many steps, so watch this chart for automations that loop or retry.

## Investigate unexpected usage

1. Narrow the date range to when usage rose.
2. Match it with recent deployments, imports, traffic or automation runs.
3. Use [Logs](/docs/dashboard/logs) for request-level detail, and [run history](/docs/automations/runs) for automation outcomes. Failing work that retries still uses resources.

To control what usage can cost:

- [Spend management](/docs/spend-management) limits a Pro project's on-demand usage.
- [Auto-reload](/docs/billing/ai-credits#auto-reload) settings limit AI credit top-ups.
- [Notifications](/docs/notifications) explains the alerts Flexkit sends as a project nears its limits.


---

[View full sitemap](/docs/sitemap.md)