# How credits work

URL: https://testmode.ai/docs/how-credits-work/

> Credits pay for Testmode runs, one per step you wrote. They are reserved when a run starts and returned if a test ends before the agent acts.

Testmode runs are paid for with your organization's credits. A run costs one credit for each step written in the test cases it runs, with at least one credit per test case. Testmode reserves the full cost when the run starts, and gives credits back for any test that ends before the agent takes its first action.

## Credits belong to the organization

- **One balance for everyone.** Every project and every member of the organization draws on the same credits. Whoever starts a run, the credits come from the organization.
- **Everyone can see it.** The balance shows under the organization's name at the bottom of the sidebar, as "N credits". It turns amber when the balance runs low.
- **Owners and Admins buy credits.** They also see the balance in detail and the history of every credit on the Billing tab. See [Buy credits and read billing history](https://testmode.ai/docs/credits-and-billing/).

Starting a run, and so spending credits, needs the Editor role or above.

## What a run costs

A run costs one credit per step in each test case it runs, with a minimum of one credit per test case. The run's cost is the sum over the test cases it actually runs: the enabled ones with steps or an Expected Outcome.

Here is what a run of Pellbrook Digital's **Storefront smoke** plan costs:

| Test case | Steps | Credits |
| --- | --- | --- |
| Search finds hammers | 2 | 2 |
| Filter catalogue by the Hammer category | 2 | 2 |
| Product page shows price and Add to cart | 2 | 2 |
| Add a product to the cart | 4 | 4 |
| Login rejects an unknown email | 2 | 2 |
| **The whole run** | | **12** |

Good to know:

- **Testmode counts the steps you wrote, not the browser actions.** One step can take the agent several clicks and keystrokes. It still costs one credit.
- **Disabled test cases cost nothing.** They are left out of the run.
- **Test cases with no steps and no Expected Outcome are skipped and cost nothing.** The **Run tests** dialog marks them and leaves them out of its estimate. See [Run tests](https://testmode.ai/docs/run-tests/#test-cases-without-steps-are-skipped).
- **You see the cost before you start.** The **Run tests** dialog shows "One credit per test step" until you choose what to run, then "Uses N credits · M available". Testmode confirms the cost when it creates the run.
- **Re-run costs the same way.** **Re-run** on the Runs page starts a new run and uses credits, but it does not show the cost first.
- **Plans and prices are on the [Pricing](https://testmode.ai/pricing/) page.** The test-run counts there are estimates, because the credits a run uses depend on how many steps its tests have.

**Note: About the wording on the Billing tab**

The Billing tab describes the charge as one credit per executed test step. The charge is actually worked out from the steps written in each test case and reserved in full when the run starts, however many actions the agent takes.

## When credits are reserved

Testmode reserves the whole cost of a run at the moment it creates the run, before any test starts. The credits leave the balance straight away.

- **All or nothing.** If the balance is smaller than the cost, no run is created. There are no partial runs.
- **The cost is fixed at the start.** Editing a test case's steps after the run has started does not change what that run costs.
- **A test stays charged once the agent has acted.** When the agent recorded at least one action, the test keeps its credits, whether it passed, failed, ran out of time or hit an error. Finding what broke is what the run is for.
- **A verdict without actions counts too.** When the agent reaches its verdict just by looking at the page, Testmode records an "Observed the page" step and the test keeps its credits.
- **A run cannot be stopped once it starts.** Its credits are spent, apart from those returned as described in the next section. See [Run tests](https://testmode.ai/docs/run-tests/#stop-a-run).

## Credits returned for tests that never ran

Testmode gives a test's credits back when the agent never took an action in it. That happens when:

- its test case was deleted before the test started
- its environment was deleted before the test started
- it had nothing left to check when it started, for example because its steps were blank once variables were filled in. It fails with "This test has no steps or expected outcome to check. Add at least one step."
- automatic login failed before the test started ("Login failed: …")
- it hit an error or ran out of time before the agent took its first action
- it never started, even after Testmode tried several times

The History on the Billing tab then shows "Refund (test did not run)", linked to the run. The test itself still shows `Failed` on the run page, with the reason, and its **Test Steps** read "No Steps Recorded". See [Troubleshoot login](https://testmode.ai/docs/troubleshoot-login/) when the reason is a login failure.

## When there are not enough credits

**Runs you start.** When the cost is higher than the balance, the footer of the **Run tests** dialog turns amber and **Run** is disabled. Owners and Admins see a **Buy credits** link that opens Billing. Everyone else sees "ask an organization admin to buy credits".

[Screenshot: Run tests dialog for a test plan on Production, with an amber footer saying the run uses more credits than are available and asking an organization admin to buy credits, and Run disabled. Caption: When the organization does not have enough credits, Run is disabled and the footer says why.]

If the balance drops after you opened the dialog, Testmode refuses the run when you click **Run** and shows "Not enough credits", with how many credits the run needs and how many the organization has.

**Scheduled runs.** When a schedule fires without enough credits, no tests run. Testmode records a run with status `Skipped` and the reason "Insufficient credits: this run needs X credits but the organization has Y." A scheduled run is also skipped, at no cost, when none of its test cases has steps or an Expected Outcome.

- A skipped run is not charged, and it sends no email report.
- Its page shows "This run was skipped" with a **Manage credits** link.
- The next scheduled time tries again. Skipped runs are not retried.

See [Schedule test plans](https://testmode.ai/docs/schedule-test-plans/).

**Explorations.** Starting an exploration needs a credit balance above zero. With no credits left, Testmode refuses with "Your organization has no credits left. Buy credits to start an exploration." See [Start an exploration](https://testmode.ai/docs/start-an-exploration/#before-you-start).

## Expiry

Credits are added in batches, and every batch has an expiry date. Before you buy, a credit pack can show how long its credits stay valid, such as "Valid for N days".

- **Soonest first.** Testmode spends credits from the batch that expires soonest.
- **Expired credits stop counting at once.** Unused credits leave the balance when their batch expires, and the History later shows "Credits expired".
- **Next expiry** on the Billing tab shows how many credits expire next, and when.

## Where to see your balance

| Where | Who sees it | What it shows |
| --- | --- | --- |
| The organization button at the bottom of the sidebar | Every member | The balance, in amber when it runs low. |
| The footer of the **Run tests** dialog | Editors and above, when starting a run | What the run will use and what is available. |
| **Billing & credits** in the organization menu | Owners and Admins | The balance, **Next expiry**, **Used in the last 30 days** and the full History. |

When the balance runs low, the Billing tab reads "Running low. Buy credits to keep tests running." At zero it reads "No credits left. Test runs are paused until you buy more."

## Related

  - [Buy credits and read billing history](https://testmode.ai/docs/credits-and-billing/): Buy credit packs, open invoices and read the credit history.
  - [Run tests](https://testmode.ai/docs/run-tests/): Check the cost in the Run tests dialog and start a run.
  - [Run and test statuses](https://testmode.ai/docs/run-and-test-statuses/): What Skipped, Failed and the other statuses mean.