> ## Documentation Index
> Fetch the complete documentation index at: https://www.voiceflow.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Response checks

> A Response check evaluates an agent turn. Choose the judge in the check’s Judge using controls. LLM as judge is useful when meaning, tone, or required information…

A **Response** check evaluates an agent turn. Choose the judge in the check’s **Judge using** controls.

## Settings

| Setting | Values | Default | Effect |
| --- | --- | --- | --- |
| Judge using | **LLM as judge**, **Exact response** | LLM as judge for a new response check | Chooses descriptive evaluation or an exact text match. |
| Describe expected response | Text | Empty | The expectation used by LLM as judge. |
| Exact response | Text | Empty when switching to this mode | The exact, case-sensitive response required to pass. |

## Behavior and dependencies

LLM as judge is useful when meaning, tone, or required information matters more than one wording. Exact response is useful for a response that must match a specific string.

Changing the judge creates a new empty expectation for that mode. Re-enter the expected text after changing modes. The editor explicitly states that an empty LLM expectation passes when the agent does not respond; use a meaningful expectation when a reply is required.

## Example

**Illustrative LLM expectation:** “Explain that the customer needs an order number before checking status.”

**Illustrative exact response:** `Please provide your order number.`

The first permits different wording. The second checks the specified text.

## Related settings

- [Independent and sequential checks](/docs/platform/rehearsal/tests/execution-order)
- [Routing checks](/docs/platform/rehearsal/tests/routing-checks)
- [Review results and failures](/docs/platform/rehearsal/runs/failures)
