---
title: "Enable auto-apply | Grafana Cloud documentation"
description: "Enable auto-apply Auto-apply automatically applies recommendations for you, eliminating manual effort and enabling your environment to be continuously fine-tuned for optimal cost and observability outcomes."
---

> For a curated documentation index, see [llms.txt](/llms.txt). For the complete documentation index, see [llms-full.txt](/llms-full.txt).

# Enable auto-apply

Auto-apply automatically applies recommendations for you, eliminating manual effort and enabling your environment to be continuously fine-tuned for optimal cost and observability outcomes.

## Before you begin

To enable auto-apply, you need the **Admin** role.

Create [exemptions](/docs/grafana-cloud/adaptive-telemetry/adaptive-metrics/create-exemptions/) before you turn on auto-apply, so critical metrics aren’t aggregated inadvertently.

For more information, refer to [Manage access to Grafana Adaptive Metrics](/docs/grafana-cloud/adaptive-telemetry/adaptive-metrics/additional-configuration/adaptive-metrics-rbac/).

## Auto-apply workflow

Enable auto-apply using the UI by updating the default segment. As a result, rule recommendations are applied without requiring any manual review or approval.

If you are using segmentation, you can enable auto-apply on a per-segment basis or edit existing segments to enable auto-apply.

1. Update the default segment to enable auto-apply.
   
   If using segmentation, you can add a segment or edit an existing segment to enable auto-apply.
2. Select **Coverage-first** or **Cost-first** for the segment.
3. The system analyzes the metrics within the specific segment and generates optimization recommendations based on the analysis.
4. The system automatically applies the recommendations allowed by the selected mode.

> Note
> 
> You can still add and edit custom rules after you enable auto-apply.
> 
> Auto-apply doesn’t modify affix (prefix or suffix) aggregation rules or drop rules. Auto-apply can remove or replace exact-match aggregation rules. To keep an exact-match rule in place, [create an exemption](/docs/grafana-cloud/adaptive-telemetry/adaptive-metrics/create-exemptions/) for that metric.

## Enable auto-apply using the UI

Enable auto-apply through the Grafana user interface, which provides you with a more visual way of managing auto-apply. It is the recommended way for initial set up and smaller organizations and it is also easier to make free-form adjustments.

### Choose an auto-apply mode

You can select an auto-apply mode for each segment. The mode controls which recommendations are applied automatically.

You choose a mode when you enable auto-apply for the Default segment or for multiple segments on the **Overview** page, for a single segment on the **Segments** page, and in the prompt after **Apply all recommendations**.

**Coverage-first** applies every recommendation, including any that are expected to add series when new usages of previously aggregated data need coverage. Recommended rule changes can increase the series count, for example, when a new query or dashboard usage requires more raw data than was previously available. Use this mode when adapting automatically to observed metric usage is more important than limiting estimated series growth.

**Cost-first** applies only those recommendations that are expected to reduce or maintain the overall series count. This mode uses estimated savings from the most recent analysis of your metric and usage data. It applies all recommendations that are expected to reduce your series count. It also applies all recommendations that are expected to increase your series count, but only up to the budget created by series count reductions applied in the same cycle. Recommendations that exceed this budget aren’t applied automatically. You can still apply those recommendations from the UI. Use this mode when limiting estimated series growth is more important than automatically expanding coverage.

Both modes are subject to the limit on aggregation rules. If a segment is at or near its limit, auto-apply might skip some or all recommendations.

> Caution
> 
> Cost-first mode makes decisions using the series count estimates generated during the last analysis cycle. It doesn’t compare the live series count before and after applying rule changes. Cost-first mode is a guardrail based on recent series count estimates, not a hard guarantee that series count never increases under any circumstances.

### Enable auto-apply for a single segment

Start with the Default segment if you are not using segmentation yet. The Default segment holds the series that are not assigned to a named segment. On **Segments**, you can also enable auto-apply for one existing segment, or while you add a segment.

1. Navigate to **Adaptive Metrics**.
2. On the **Overview** page, click **Enable auto-apply** next to the **Segment** dropdown.
   
   The **Enable auto-apply per segment** dialog box opens.
3. Select **Default**.
   
   If you only have the Default segment, the dialog box recommends creating more segments and includes a **Go to Segments** link. You can still enable auto-apply for Default.
4. Select an **Auto-apply mode**.
5. Click **Apply**.

To enable auto-apply for one segment from **Segments**, including a segment you are adding:

1. Navigate to the **Segments** page.
2. Click **Edit segment**, or **Edit default segment** for the Default segment. To turn auto-apply on while you create a segment, click **Add new segment** and fill out the segment details.
3. Select the **Enable auto-apply** checkbox.
4. Select an **Auto-apply mode**.
5. Click **Apply** for the Default segment, **Save** for a named segment, or **Add** when you are creating the segment.
   
   In the Segments list, you can view the segment and whether auto-apply is enabled. From the **Metrics** page, you can search for a segment and see whether auto-apply is enabled.

### Enable auto-apply for multiple segments

Use this when you already have segments and want one mode for every segment you select. The mode you select applies to every segment in the selection.

1. Navigate to **Adaptive Metrics**.
2. On the **Overview** page, click **Enable auto-apply** next to the **Segment** dropdown.
   
   The **Enable auto-apply per segment** dialog box opens.
3. Select the segments you want to enable auto-apply for.
   
   Segments that already have auto-apply enabled aren’t included.
4. Select an **Auto-apply mode**.
5. Click **Apply**.

### Enable auto-apply after applying all recommendations

After you click **Apply all recommendations** on the **Metrics** page and click **Apply** in the **Change rules?** dialog box, a prompt asks you to enable auto-apply for the current segment.

The prompt doesn’t appear if you apply a selection of recommendations, or if auto-apply is already on for that segment.

Select an **Auto-apply mode** in the prompt. To change the mode later, on the **Segments** page click **Edit segment** or **Edit default segment**. The **Overview** dialog doesn’t list a segment that already has auto-apply on.

1. Select an **Auto-apply mode**.
2. Click **Enable auto-apply for this segment** to turn on auto-apply, or **Dismiss** to close the prompt.
3. Optionally, select **Do not show this again** before you dismiss the prompt if you don’t want to see it later.

### Turn auto-apply off

1. Navigate to **Adaptive Metrics**.
2. Navigate to the **Segments** page.
3. Click **Edit segment**, or **Edit default segment** for the Default segment.
4. Clear **Enable auto-apply**.
5. Click **Apply** for the Default segment, or **Save** for a named segment.

The rules stay as they were applied at that moment. You can still change them yourself.

After auto-apply is on, return to **Overview**. **Time series reduction** shows the reduction from applied rules. On later visits, the recommendations summary shows how many recommendations are still pending. With **Cost-first**, recommendations past the budget stay pending. Refer to [Monitor and validate results](/docs/grafana-cloud/adaptive-telemetry/adaptive-metrics/monitor/).

## Enable auto-apply using the API

Enable auto-apply using the API. This is useful for automation, version control and scalability for managing a large number of teams and integrating with existing as-code practices.

The auto-apply feature can be turned on and off on a per segment basis. This allows you to turn it on for segments where you are confident that you want to apply recommendations and leave it turned off for other segments.

Set `auto_apply.gate.policy` to one of the following values:

- `unbounded` selects Coverage-first.
- `no-increase` selects Cost-first.

If you omit `gate` or `policy`, auto-apply uses Coverage-first.

The `default` segment always exists, even without a custom segment and is configured using the main recommendations configuration.

### Configure the default segment

1. Use the `/aggregations/recommendations/config` endpoint.
2. Set the `enabled` property to true.
3. Set `gate.policy` to the policy for the mode you want to use.

The following example enables Cost-first.

![Copy code to clipboard](/media/images/icons/icon-copy-small-2.svg) Copy

```none
# Get the current config
% curl -s -u '<user>:<key>' 'https://<prom URL>/aggregations/recommendations/config' | jq '.'
{
  "keep_labels": [],
  "auto_apply": {
    "enabled": false
  }
}

# Enable auto-apply in Cost-first mode
curl -s -u '<user>:<key>' 'https://<prom URL>/aggregations/recommendations/config' -X POST -d '{
  "keep_labels": [],
  "auto_apply": {
    "enabled": true,
    "gate": {
      "policy": "no-increase"
    }
  }
}'

# Verify the updated setting
% curl -s -u '<user>:<key>' 'https://<prom URL>/aggregations/recommendations/config' | jq '.'
{
  "keep_labels": [],
  "auto_apply": {
    "enabled": true,
    "gate": {
      "policy": "no-increase"
    }
  }
}
```

### Configure custom segments

Configure auto-apply on a per-segment basis using the segment configuration endpoint.

1. Go to the `/aggregations/rules/segments` endpoint.
2. Set the `enabled` property to true.
3. Set `gate.policy` to the policy for the mode you want to use.

The following example enables Cost-first.

![Copy code to clipboard](/media/images/icons/icon-copy-small-2.svg) Copy

```none
# Get the current config to look up the segment ID
% curl -s -u '<user>:<key>' 'https://<prom URL>/aggregations/rules/segments' | jq '.'
[
  {
    "id": "01JTRP8G515JA77KD7K7ABFEKQ",
    "name": "Team1",
    "selector": "{team=\"team1\"}",
    "auto_apply": {
      "enabled": false
    }
  }
]

# Enable auto-apply in Cost-first mode on the segment with ID 01JTRP8G515JA77KD7K7ABFEKQ
% curl -s -u '<user>:<key>' 'https://<prom URL>/aggregations/rules/segments?segment=01JTRP8G515JA77KD7K7ABFEKQ' -X PUT -d '{
  "id": "01JTRP8G515JA77KD7K7ABFEKQ",
  "name": "Team1",
  "selector": "{team=\"team1\"}",
  "auto_apply": {
    "enabled": true,
    "gate": {
      "policy": "no-increase"
    }
  }
}'

# Verify the updated setting
% curl -s -u '<user>:<key>' 'https://<prom URL>/aggregations/rules/segments' | jq '.'
[
  {
    "id": "01JTRP8G515JA77KD7K7ABFEKQ",
    "name": "Team1",
    "selector": "{team=\"team1\"}",
    "auto_apply": {
      "enabled": true,
      "gate": {
        "policy": "no-increase"
      }
    }
  }
]
```

## Enable auto-apply using Terraform

Manage auto-apply as part of your infrastructure-as-code strategy using the `grafana-adaptive-metrics` Terraform provider.

Use the [`grafana-adaptive-metrics_recommendations_config` resource](https://registry.terraform.io/providers/grafana/grafana-adaptive-metrics/latest/docs/resources/recommendations_config) to configure the default segment. Use the [`grafana-adaptive-metrics_segment` resource](https://registry.terraform.io/providers/grafana/grafana-adaptive-metrics/latest/docs/resources/segment) to configure custom segments.

Set `auto_apply.enabled` to `true` for each segment where you want auto-apply on.

Set `auto_apply.gate.policy` to `unbounded` for Coverage-first or `no-increase` for Cost-first.

For provider setup instructions, refer to [Adaptive Metrics Terraform provider](/docs/grafana-cloud/adaptive-telemetry/adaptive-metrics/manage-as-code/adaptive-metrics-terraform-provider/).