---
title: AI connector
description: Documentation for the deprecated AI Connector, used to send requests to LLM services from Kibana through the Elasticsearch Inference API.
url: https://www.elastic.co/elastic/docs-builder/docs/4665/reference/kibana/connectors-kibana/ai-connector
products:
  - Kibana
applies_to:
  - Elastic Cloud Serverless: Deprecated
  - Elastic Stack: Deprecated since 9.5, Preview from 9.2 to 9.4
---

# AI connector
<important>
  This connector is being replaced by Elasticsearch inference endpoints and will be removed entirely in Elastic Stack 10.0 or in January 2028, whichever comes first. To configure access to an LLM provider, create an inference endpoint with the `chat_completion` task type. For steps that use your own provider account and API key, refer to [External inference](https://docs-v3-preview.elastic.dev/elastic/docs-builder/docs/4665/explore-analyze/elastic-inference/external). To use models that Elastic hosts instead, refer to [Elastic Managed LLMs](https://www.elastic.co/elastic/docs-builder/docs/4665/reference/kibana/connectors-kibana/elastic-managed-llm).Availability of this connector depends on your deployment and version:
  - <applies-to>Elastic Cloud Serverless: Deprecated</applies-to> <applies-to>Elastic Stack: Planned for deprecation</applies-to> You can no longer create new connectors of this type by selecting **Create connector** in Kibana. Existing connectors and their rule actions continue to work, but the **Connectors** page hides them by default. To show them, open **Display options** and set **Deprecated** to **Show**.
  - <applies-to>Elastic Stack: Deprecated in 9.5</applies-to> This connector is still available, but deprecated.
</important>

The AI Connector uses the [Inference API](https://docs-v3-preview.elastic.dev/elastic/docs-builder/docs/4665/explore-analyze/elastic-inference/inference-api) to connect to third-party Large Language Model (LLM) services to enable AI-powered Kibana features such as AI Assistant. It can connect with LLM services including Amazon Bedrock, Azure, Google Gemini, and OpenAI. It can also connect with models you've enabled in Elastic Inference Service (EIS).

## Requirements

- A working [LLM connector](https://docs-v3-preview.elastic.dev/elastic/docs-builder/docs/4665/explore-analyze/ai-features/llm-guides/llm-connectors).
- Elastic Stack users: An [Enterprise](https://www.elastic.co/pricing) subscription.
- Serverless users: a project with the [Security Analytics Complete](https://docs-v3-preview.elastic.dev/elastic/docs-builder/docs/4665/deploy-manage/deploy/elastic-cloud/project-settings) feature tier.
- The **Actions and Connectors : All** [privilege](https://docs-v3-preview.elastic.dev/elastic/docs-builder/docs/4665/deploy-manage/users-roles/cluster-or-deployment-auth/kibana-privileges).


## Set up an AI connector

<applies-to>
  - Elastic Cloud Serverless: Unavailable
  - Elastic Stack: Planned for removal
  - Elastic Stack: Deprecated in 9.5
  - Elastic Stack: Preview from 9.2 to 9.4
</applies-to>

1. Use the [global search field](https://docs-v3-preview.elastic.dev/elastic/docs-builder/docs/4665/explore-analyze/find-and-organize/find-apps-and-objects) to find the **Connectors** page.
2. Click **+ Create connector**. Select **AI Connector**.
3. Name your connector and select which LLM service to use.
4. In the **Settings** section, specify which model to use for this connector. You may need to refer to your selected LLM service's list of models. Model names and other model parameters must be exact.
5. In the **Authentication** section, provide the necessary credentials. All supported LLM services other than EIS require authentication. For information about how to generate authentication information for each one, refer to [Enable LLM access](https://docs-v3-preview.elastic.dev/elastic/docs-builder/docs/4665/explore-analyze/ai-features/llm-guides/llm-connectors).
6. Expand the **Additional settings** menu. Select a **Task type** that matches your selected model's purpose. Refer to the model information for your selected model to identify which task types it supports.
7. (Optional) In the **Additional settings** section, some LLM services allow you to set a **Context window length**. Enter a value to specify how many tokens to send to the model per request, or leave it blank to use the model's default.
8. Finally, click **Save & test** to verify that the connector is set up.

You can now select your newly configured connector for all Kibana features that rely on LLM connectors.