---
updatedAt: 2025-12-08T15:53:56.000Z
agentTools:
  projectIndex: https://docs.arthur.ai/llms.txt
---

# Setting Up GenAI Metrics

Learn how to configure built-in and custom metrics that evaluate your GenAI model’s prompts and outputs for safety, compliance, and quality.

## Accessing the Model Metrics Page

Navigate to **Model Overview → Model Management → Metrics**.

<Image alt="Metrics Navigation" border={false} src="https://files.readme.io/8b5a0df5ea0b48ab764078f996cb2446fe71eb654ec5bfdf818b98c1bd4dd450-Screenshot_2025-07-10_at_12.08.40.png" />

This opens the **Model Metrics** configuration screen:

<Image alt="Metrics Screen" border={false} src="https://files.readme.io/ed8d36f3f85d3c006b145084fda29908b5e78a515098a805a7706578e83012fa-Screenshot_2025-05-06_at_11.30.34.png" />

If your model includes default rules, they will appear but cannot be modified here:

<Image alt="Default Rules" border={false} src="https://files.readme.io/9cfaa65691b99612142da7aa3d70f5bf47d47d77076c5a226ec7b370e85fbd28-Screenshot_2025-05-06_at_11.32.54.png" />

***

## Adding a New Metric

Click **+ Metric** to open the metric creation form:

<Image alt="Add Metric Form" border={false} src="https://files.readme.io/c6d62261e1a54f11da276a5594b5214709b48ba5d57de16844aa35e236c81df8-Screenshot_2025-05-06_at_11.35.53.png" />

You must provide:

1. **Metric Name**
2. **Metric Type**

Supported metric types include:

* Toxicity
* PII
* Sensitive Data
* Prompt Injection
* Hallucination
* Keyword
* Regex

***

# Metric Types & Configuration

## Toxicity

Set a threshold using the slider or manual value input.

<Image alt="Toxicity Config" border={false} src="https://files.readme.io/c27cb0b44bad3afa08ade60fd0839ede512bdcea22404190555229cf5b6d0ddf-Screenshot_2025-05-06_at_11.49.28.png" />

**Applies to:** Prompt, Response, or Both.

***

## PII

Choose which PII entities the metric should detect.

Loom demo:

<https://www.loom.com/embed/22f017056bad46a8a0d64eea1ed157f6>

**Applies to:** Prompt and Response.

***

## Sensitive Data

Add custom sensitive terms and flag them accordingly.

Loom demo:\
<https://www.loom.com/embed/56575adfd48d4ccab9d9b54a496abf90>

**Applies to:** Prompt only.

***

## Prompt Injection

Detects attempts to manipulate or override system instructions.

<Image alt="Prompt Injection" border={false} src="https://files.readme.io/605d74c4c422e5a199d2d75f84138db3643dc0e9ce08c3b51e84595853a13492-Screenshot_2025-05-06_at_12.30.19.png" />

**Applies to:** Prompt only.

***

## Hallucination

Detects unsupported or fabricated model responses.

<Image alt="Hallucination" border={false} src="https://files.readme.io/346e525d20762d3882ef396ac4f0de126f58b8612501691c63ea703bce33c4c1-Screenshot_2025-05-06_at_12.43.47.png" />

**Applies to:** Response only.

***

## Keyword

Provide a list of keywords you want to detect.

<Image alt="Keyword" border={false} src="https://files.readme.io/8f9c47d5d35dba222e183ac418e034a5eeaca09c5daafa207dd739859293f90f-Screenshot_2025-05-06_at_13.04.51.png" />

**Applies to:** Prompt, Response, or Both.

***

## Regex

Define regex-based pattern detection.

<Image alt="Regex" border={false} src="https://files.readme.io/8e009aebe3065f2dd016afc50330280f8b154ece1bbedaf266b8136abcd9d97b-Screenshot_2025-05-06_at_13.09.41.png" />

**Applies to:** Prompt, Response, or Both.

***

# Reviewing & Managing Metrics

All newly added metrics appear in the list:

Loom walkthrough:\
<https://www.loom.com/embed/dcf4521aa41e4bf9bbfb94029fe53d9d>

New metrics remain enabled until saved:

<Image alt="Enabled Metrics" border={false} src="https://files.readme.io/20ab58825ded9117126efd043c1ea5487b85666314db240fb4e6555ae0a91bb0-Screenshot_2025-05-06_at_13.16.32.png" />

You may edit or remove metrics at any time:

Loom demo:\
<https://www.loom.com/embed/7019ea04b7114b4f8f152a2bd1a42930>

Once everything is configured, click **Save**. After saving, each metric can be toggled on/off as needed:

<Image alt="Save Metrics" border={false} src="https://files.readme.io/502924573dd5dafa19633bc22a9f866017414c047e057e24b5f762326d8f3ce3-Screenshot_2025-05-06_at_13.22.40.png" />

***

# What’s Next

* Run evaluation pipelines using the metrics you've configured.
* Enable real-time guardrails to enforce these checks at inference time.
* Explore additional guides:
  * Toxicity Detection
  * Sensitive Data Leakage
  * Prompt Injection Detection
  * Hallucination Detection