diff --git a/_api-reference/asynchronous-search/async-search-stats.md b/_api-reference/asynchronous-search/async-search-stats.md
new file mode 100644
index 0000000000..80f3d557e6
--- /dev/null
+++ b/_api-reference/asynchronous-search/async-search-stats.md
@@ -0,0 +1,126 @@
+---
+layout: default
+title: Asynchronous search stats
+parent: Asynchronous Search APIs
+nav_order: 40
+---
+
+# Asynchronous search stats
+**Introduced 1.0**
+{: .label .label-purple }
+
+
+The Asynchronous Search Stats API provides statistics about asynchronous search activity across the cluster. Use this API to monitor performance, track usage trends, and evaluate the resource impact of running and completed asynchronous searches.
+
+The following list describes common use cases for this API:
+
+- Monitoring and tuning performance.
+- Troubleshooting search issues.
+- Planning cluster capacity.
+- Understanding how asynchronous search is used.
+
+<!-- spec_insert_start
+api: asynchronous_search.get
+component: endpoints
+-->
+## Endpoints
+```json
+GET /_plugins/_asynchronous_search/{id}
+```
+<!-- spec_insert_end -->
+
+## Example request
+
+The following request retrieves statistics about asynchronous searches across all nodes in the cluster:
+
+```json
+GET /_plugins/_asynchronous_search/stats
+```
+{% include copy-curl.html %}
+
+## Example response
+
+```json
+{
+  "_nodes": {
+    "total": 2,
+    "successful": 2,
+    "failed": 0
+  },
+  "cluster_name": "opensearch-cluster",
+  "nodes": {
+    "node1": {
+      "running_current": 0,
+      "running_count": 12,
+      "persisted_current": 3,
+      "persisted_count": 15,
+      "completed_count": 25,
+      "cancelled_count": 2,
+      "rejected_count": 1,
+      "failed_count": 3
+    },
+    "node2": {
+      "running_current": 1,
+      "running_count": 8,
+      "persisted_current": 2,
+      "persisted_count": 10,
+      "completed_count": 18,
+      "cancelled_count": 1,
+      "rejected_count": 0,
+      "failed_count": 2
+    }
+  }
+}
+```
+
+## Response body fields
+
+The response body is a JSON object with the following fields.
+
+| Property | Data type | Description |
+| :--- | :--- | :--- |
+| `_nodes` | Object | Statistics about node operations and responses. |
+| `cluster_name` | String | The name of the OpenSearch cluster. |
+| `nodes` | Object | A map of node IDs to node-specific asynchronous search statistics. |
+
+<details markdown="block">
+  <summary>
+    Response body fields: <code>_nodes</code>
+  </summary>
+  {: .text-delta}
+
+The statistics about node operations.
+
+`_nodes` is a JSON object with the following fields.
+
+| Property | Required | Data type | Description |
+| :--- | :--- | :--- | :--- |
+| `failed` | **Required** | Integer | The number of nodes that rejected the request or failed to respond. If this value is not 0, then a reason for the rejection or failure is included in the response. |
+| `successful` | **Required** | Integer | The number of nodes that responded successfully to the request. |
+| `total` | **Required** | Integer | The total number of nodes selected by the request. |
+| `failures` | Optional | Array of Objects | Details about any failures that occurred during the request. Present only when there are failures. |
+
+</details>
+
+<details markdown="block">
+  <summary>
+    Response body fields: <code>nodes</code>
+  </summary>
+  {: .text-delta}
+
+A map of node IDs to node-specific asynchronous search statistics.
+
+Each entry in the `nodes` object contains the following fields:
+
+| Property | Data type | Description |
+| :--- | :--- | :--- |
+| `running_current` | Integer | The number of asynchronous searches currently running on this node. |
+| `running_count` | Integer | The total number of asynchronous searches that have been executed on this node since it was started. |
+| `persisted_current` | Integer | The number of completed asynchronous searches that are currently persisted on this node. |
+| `persisted_count` | Integer | The total number of asynchronous searches that have been persisted on this node since it was started. |
+| `completed_count` | Integer | The total number of asynchronous searches that have completed successfully on this node since it was started. |
+| `cancelled_count` | Integer | The total number of asynchronous searches that have been cancelled on this node since it was started. |
+| `rejected_count` | Integer | The total number of asynchronous searches that have been rejected on this node since it was started. |
+| `failed_count` | Integer | The total number of asynchronous searches that have failed on this node since it was started. |
+
+</details>
\ No newline at end of file
diff --git a/_api-reference/asynchronous-search/create-async-search.md b/_api-reference/asynchronous-search/create-async-search.md
new file mode 100644
index 0000000000..c46c10f0e3
--- /dev/null
+++ b/_api-reference/asynchronous-search/create-async-search.md
@@ -0,0 +1,242 @@
+---
+layout: default
+title: Create asynchronous search
+parent: Asynchronous Search APIs
+nav_order: 10
+---
+
+# Create asynchronous search
+**Introduced 1.0**
+{: .label .label-purple }
+
+
+The Create Asynchronous Search API lets you run a search in the background and retrieve results later. This is useful for large or complex queries that might time out if executed synchronously. Asynchronous searches perform the following actions:
+
+- Run in the background.
+- Return a search ID immediately.
+- Let you check the status or get partial results while the search is running.
+- Allow access to full results once complete.
+
+You can configure how long the search results are stored in the cluster after completion.
+
+
+<!-- spec_insert_start
+api: asynchronous_search.search
+component: endpoints
+-->
+## Endpoints
+```json
+POST /_plugins/_asynchronous_search
+```
+<!-- spec_insert_end -->
+
+<!-- spec_insert_start
+api: asynchronous_search.search
+component: query_parameters
+-->
+## Query parameters
+
+The following table lists the available query parameters. All query parameters are optional.
+
+| Parameter | Data type | Description |
+| :--- | :--- | :--- |
+| `index` | String | The name of the index to be searched. Can be an individual name, a comma-separated list of indexes, or a wildcard expression of index names. |
+| `keep_alive` | String | The amount of time that the result is saved in the cluster. For example, `2d` means that the results are stored in the cluster for 48 hours. The saved search results are deleted after this period or if the search is canceled. Note that this includes the query execution time. If the query exceeds this amount of time, the process cancels this query automatically. |
+| `keep_on_completion` | Boolean | Whether to save the results in the cluster after the search is complete. You can examine the stored results at a later time. |
+| `wait_for_completion_timeout` | String | The amount of time to wait for the results. You can poll the remaining results based on an ID. The maximum value is 300 seconds. Default is `1s`. |
+
+<!-- spec_insert_end -->
+
+## Request body fields
+
+## Request body fields
+
+The request body follows the standard [Search API format]({{site.url}}{{site.baseurl}}/query-dsl/) using the Query DSL. The body is optional.
+
+| Field | Data type | Description |
+| :--- | :--- | :--- |
+| `_source` | All | Controls how the `_source` field is returned. You can disable fetching or filter specific fields. |
+| `aggregations` | Object | Defines aggregation operations to perform during the search. |
+| `collapse` | Object | Collapses search results based on a field value. |
+| `docvalue_fields` | Array | Returns doc values for matching fields in the `hits.fields` section. Accepts wildcards. |
+| `explain` | Boolean | When `true`, includes score explanation for each hit. |
+| `ext` | Object | Plugin-specific search extensions. |
+| `fields` | Array | Returns values for matching fields in `hits.fields`. Accepts wildcards. |
+| `from` | Integer | Offset for the first result to return. Used for pagination. |
+| `highlight` | Object | Configures field highlighting in the response. |
+| `indices_boost` | Array | Boosts relevance scores for specified indices. |
+| `min_score` | Float | Excludes documents with a `_score` below this threshold. |
+| `pit` | Object | Defines a point-in-time (PIT) context for consistent snapshots. |
+| `post_filter` | Object | Applies a filter after aggregations are computed. |
+| `profile` | Boolean | When `true`, includes profiling data for debugging. |
+| `query` | Object | Defines the search criteria using the Query DSL. |
+| `rank` | Object | Configures ranking algorithms or features. |
+| `script_fields` | Object | Returns custom script-based values for each hit. |
+| `search_after` | Array | Enables pagination beyond 10,000 results using sort values. |
+| `seq_no_primary_term` | Boolean | When `true`, includes sequence number and primary term metadata. |
+| `size` | Integer | Number of hits to return. |
+| `slice` | Object | Enables parallel execution of a sliced scroll request. |
+| `sort` | Array | Sorts results by specified fields. Valid values: `_score`, `_doc`. |
+| `stats` | Array | Assigns search statistics to named groups. |
+| `stored_fields` | Array | Specifies stored fields to return in the response. |
+| `suggest` | Object | Provides term or phrase suggestions based on input text. |
+| `terminate_after` | Integer | Stops collecting results after a specified number per shard. |
+| `timeout` | String | Maximum time to wait for shard responses before failing the request. |
+| `track_scores` | Boolean | When `true`, calculates scores even if not used for sorting. |
+| `track_total_hits` | Boolean or Integer | Controls total hit count calculation. `true` for exact count; `false` disables it. |
+| `version` | Boolean | When `true`, includes the document version in hits. |
+
+
+
+## Example requests
+
+### Basic asynchronous search
+
+The following example submits an asynchronous search for the term "OpenSearch" across all indices with a 2-minute wait time:
+
+```json
+POST /_plugins/_asynchronous_search?wait_for_completion_timeout=2m&keep_alive=10m&keep_on_completion=true
+{
+  "query": {
+    "match": {
+      "content": "OpenSearch"
+    }
+  },
+  "size": 100
+}
+```
+{% include copy-curl.html %}
+
+### Asynchronous search with aggregations
+
+The following example runs an asynchronous search with both a query and aggregations:
+
+```json
+POST /_plugins/_asynchronous_search?wait_for_completion_timeout=5s&keep_alive=1d
+{
+  "query": {
+    "range": {
+      "timestamp": {
+        "gte": "now-1d/d",
+        "lt": "now/d"
+      }
+    }
+  },
+  "size": 0,
+  "aggs": {
+    "categories": {
+      "terms": {
+        "field": "category",
+        "size": 10
+      }
+    }
+  }
+}
+```
+{% include copy-curl.html %}
+
+## Example responses
+
+### Response with partial results
+
+```json
+{
+  "id": "FmRldE8zREVEUzA2ZVpUeGs2ejJFUFEaMkZ5QTVrSTZSaVN3WlNFVmtlWGJKdFQ0USsydzoxMDc=",
+  "state": "RUNNING",
+  "start_time_in_millis": 1663276896192,
+  "expiration_time_in_millis": 1663363296192,
+  "response": {
+    "took": 1019,
+    "timed_out": false,
+    "_shards": {
+      "total": 10,
+      "successful": 5,
+      "skipped": 0,
+      "failed": 0
+    },
+    "hits": {
+      "total": {
+        "value": 3,
+        "relation": "eq"
+      },
+      "max_score": 0.9530773,
+      "hits": [
+        {
+          "_index": "my-index",
+          "_id": "1",
+          "_score": 0.9530773,
+          "_source": {
+            "content": "OpenSearch is a community-driven open source search and analytics suite"
+          }
+        },
+        // Additional hits omitted for brevity
+      ]
+    }
+  }
+}
+```
+
+### Response for a completed search
+
+```json
+{
+  "id": "MkZ5QTVrSTZSaVN3WlNFVmtlWGJKdFQ0USsydzoxFmRldE8zREVEUzA2ZVpUeGs2ejJFUFE=",
+  "state": "SUCCEEDED",
+  "start_time_in_millis": 1663276892345,
+  "expiration_time_in_millis": 1663363292345,
+  "took": 3876,
+  "response": {
+    "took": 3876,
+    "timed_out": false,
+    "_shards": {
+      "total": 10,
+      "successful": 10,
+      "skipped": 0,
+      "failed": 0
+    },
+    "hits": {
+      "total": {
+        "value": 15,
+        "relation": "eq"
+      },
+      "max_score": 2.341092,
+      "hits": [
+        // Hits omitted for brevity
+      ]
+    },
+    "aggregations": {
+      "categories": {
+        "doc_count_error_upper_bound": 0,
+        "sum_other_doc_count": 0,
+        "buckets": [
+          {
+            "key": "documentation",
+            "doc_count": 7
+          },
+          {
+            "key": "blog",
+            "doc_count": 5
+          },
+          {
+            "key": "tutorial",
+            "doc_count": 3
+          }
+        ]
+      }
+    }
+  }
+}
+```
+
+## Response body fields
+
+The response body is a JSON object with the following fields.
+
+| Property | Data type | Description |
+| :--- | :--- | :--- |
+| `expiration_time_in_millis` | Long | The UNIX timestamp in milliseconds when the asynchronous search results will expire and be deleted from the cluster. |
+| `id` | String | The unique identifier for the asynchronous search request. Use this ID to check the status, get results, or delete the search. |
+| `response` | Object | Contains the search results (if available) in the same format as a standard search response. May be empty if no results are available yet. |
+| `start_time_in_millis` | Long | The UNIX timestamp in milliseconds when the asynchronous search was started. |
+| `state` | String | The current state of the asynchronous search. Possible values include `RUNNING`, `SUCCEEDED`, `FAILED`, or `PERSISTED`. |
+| `took` | Integer | The time in milliseconds that the search took to execute so far. |
diff --git a/_api-reference/asynchronous-search/delete-async-search.md b/_api-reference/asynchronous-search/delete-async-search.md
new file mode 100644
index 0000000000..f24a2ba4bb
--- /dev/null
+++ b/_api-reference/asynchronous-search/delete-async-search.md
@@ -0,0 +1,61 @@
+---
+layout: default
+title: Delete asynchronous search
+parent: Asynchronous Search APIs
+nav_order: 30
+---
+
+# Delete asynchronous search
+**Introduced 1.0**
+{: .label .label-purple }
+
+The Delete Asynchronous Search API deletes an asynchronous search by its ID before its configured expiration time. Use this API to clean up searches that are no longer needed, free up cluster resources, or remove stored results that may contain sensitive data.
+
+The following list describes what happens when you delete an asynchronous search:
+
+- If the search is still running, the operation is canceled.
+- Any stored partial or complete results are immediately removed from the cluster.
+- Resources associated with the search are released.
+
+<!-- spec_insert_start
+api: asynchronous_search.delete
+component: endpoints
+-->
+## Endpoints
+```json
+DELETE /_plugins/_asynchronous_search/{id}
+```
+<!-- spec_insert_end -->
+
+## Path parameters
+
+The following table lists the available path parameters.
+
+| Parameter | Required | Data type | Description |
+| :--- | :--- | :--- | :--- |
+| `id` | **Required** | String | The ID of the asynchronous search to delete. This is the same ID returned when the search was created. |
+
+## Example request
+
+The following request deletes an asynchronous search with ID `FmRldE8zREVEUzA2ZV`:
+
+```json
+DELETE /_plugins/_asynchronous_search/FmRldE8zREVEUzA2ZV
+```
+{% include copy-curl.html %}
+
+## Example response
+
+```json
+{
+  "acknowledged": true
+}
+```
+
+## Response body fields
+
+The response body is a JSON object with the following fields.
+
+| Property | Data type | Description |
+| :--- | :--- | :--- |
+| `acknowledged` | Boolean | Whether the delete operation was successful. Returns `true` when the asynchronous search was successfully deleted. |
\ No newline at end of file
diff --git a/_api-reference/asynchronous-search/get-async-search.md b/_api-reference/asynchronous-search/get-async-search.md
new file mode 100644
index 0000000000..bd50165f69
--- /dev/null
+++ b/_api-reference/asynchronous-search/get-async-search.md
@@ -0,0 +1,110 @@
+---
+layout: default
+title: Get asynchronous search
+parent: Asynchronous Search APIs
+nav_order: 20
+---
+
+# Get asynchronous search
+**Introduced 1.0**
+{: .label .label-purple }
+
+The Get Asynchronous Search API retrieves the status or results of a previously submitted asynchronous search. You can use this API to:
+
+- Monitor the progress of a long-running search.
+- Retrieve partial results while the search is still running.
+- Access the final search results after completion.
+- Extend the expiration time of stored search results.
+
+This API returns both search status metadata and any available results in the same format as a standard search response.
+
+
+<!-- spec_insert_start
+api: asynchronous_search.get
+component: endpoints
+-->
+## Endpoints
+```json
+GET /_plugins/_asynchronous_search/{id}
+```
+<!-- spec_insert_end -->
+
+## Path parameters
+
+The following table lists the available path parameters.
+
+| Parameter | Required | Data type | Description |
+| :--- | :--- | :--- | :--- |
+| `id` | **Required** | String | The unique ID of the asynchronous search to retrieve. This ID is returned when the search is created. |
+
+## Query parameters
+
+The following table lists the available query parameters. All query parameters are optional.
+
+| Parameter | Data type | Description |
+| :--- | :--- | :--- |
+| `keep_alive` | String | Extends the expiration time of the search results. For example, 2d extends the expiration by two days from the current time. |
+| `wait_for_completion_timeout` | String | 	The maximum amount of time to wait for the search to complete before returning a response. Maximum is 300s. If not specified, the API returns results immediately based on the current search state. |
+
+## Example request
+
+The following request retrieves an asynchronous search with ID `FFj38GhYw82NsHg7` and extends its expiration time by one hour:
+
+```json
+GET /_plugins/_asynchronous_search/Fj38GhYw82NsHg7keep_alive=1h
+```
+{% include copy-curl.html %}
+
+## Example response
+
+```json
+{
+  "id": "Fj38GhYw82NsHg7",
+  "state": "SUCCEEDED",
+  "start_time_in_millis": 1663276896192,
+  "expiration_time_in_millis": 1663366896192,
+  "took": 3255,
+  "response": {
+    "took": 3255,
+    "timed_out": false,
+    "_shards": {
+      "total": 10,
+      "successful": 10,
+      "skipped": 0,
+      "failed": 0
+    },
+    "hits": {
+      "total": {
+        "value": 15,
+        "relation": "eq"
+      },
+      "max_score": 1.0,
+      "hits": [
+        {
+          "_index": "my-index",
+          "_id": "1",
+          "_score": 1.0,
+          "_source": {
+            "field1": "value1",
+            "field2": "value2"
+          }
+        },
+        // Additional hits omitted for brevity
+      ]
+    }
+  }
+}
+```
+
+## Response body fields
+
+The response body is a JSON object with the following fields.
+
+| Field  | Data type | Description  |
+| :---| :---- | :--- |
+| `id`     | String    | The ID of the asynchronous search.      |
+| `state`    | String    | The current state of the search. Possible values: `RUNNING`, `SUCCEEDED`, `FAILED`, or `PERSISTED`.   |
+| `start_time_in_millis`  | Long   | The time the search started, in UNIX epoch milliseconds.     |
+| `expiration_time_in_millis` | Long   | The time the results will expire, in UNIX epoch milliseconds.  |
+| `took`   | Integer   | The time the search has taken so far, in milliseconds.    |
+| `response`  | Object    | The search response object, returned in the same format as a standard search. This field may be empty if no results are available yet. |
diff --git a/_api-reference/asynchronous-search/index.md b/_api-reference/asynchronous-search/index.md
new file mode 100644
index 0000000000..fd1da6954e
--- /dev/null
+++ b/_api-reference/asynchronous-search/index.md
@@ -0,0 +1,37 @@
+---
+layout: default
+title: Asynchronous search APIs
+nav_order: 8
+has_children: true
+---
+
+# Asynchronous search
+**Introduced in 1.0**
+{: .label .label-purple }
+
+Asynchronous search in OpenSearch lets you run search operations in the background and retrieve results later. This approach is useful for long-running or resource-intensive searches that may time out in a synchronous context.
+
+## Available APIs
+
+The following APIs support asynchronous search operations:
+
+| API | Description |
+| :--- | :--- |
+| [Create asynchronous search]({{site.url}}{{site.baseurl}}/api-reference/asynchronous-search/create-async-search/) | Submits a background search request. |
+| [Get asynchronous search]({{site.url}}{{site.baseurl}}/api-reference/asynchronous-search/get-async-search/) | Retrieves the status and results of a submitted search. |
+| [Delete asynchronous search]({{site.url}}{{site.baseurl}}/api-reference/asynchronous-search/delete-async-search/) | Deletes a stored asynchronous search. |
+| [Asynchronous search stats]({{site.url}}{{site.baseurl}}/api-reference/asynchronous-search/async-search-stats/) | Returns statistics about asynchronous search usage. |
+
+## Performance considerations
+
+Asynchronous searches use cluster resources. To optimize performance, consider the following:
+
+- Set a reasonable `keep_alive` value to avoid unnecessary storage.
+- Use `keep_on_completion=false` if you plan to retrieve results immediately.
+- Monitor asynchronous search statistics to understand usage patterns.
+- Delete searches when they are no longer needed.
+- Adjust the search thread pool size if running many concurrent searches.
+
+## Security
+
+If the Security plugin is enabled, users must have either the `asynchronous_search_full_access` or `asynchronous_search_read_access` permission to manage or view asynchronous searches. You can also configure more granular permissions. For details, see [Asynchronous search security]({{site.url}}{{site.baseurl}}/search-plugins/async/security/).