Configure multimodal captioning for AI Search

task australia Updated Sep 10, 2026 2 min read configure-multimodal-captioning-for-ai-search.md

Use the AI Search Admin console to select the visual language model (VLM) provider and model for multimodal captioning.

Before you begin

The Platform Multimodal Service plugin (com.glide.platform_mm_service) must be installed on your instance. If the plugin is not installed, multimodal captioning options will not appear in the AI Search Admin console. For more information, see Activate the Platform Multimodal Service plugin.
Role required: ais_admin

About this task

When multimodal captioning is activated, attachments retrieved from supported sources are automatically analyzed and given descriptive captions, making visual content, such as images, discoverable through text-based search. You can specify the VLM provider and model to use when generating captions.

Procedure

Navigate to All > AI Search Admin > AI Search Admin Home.
Select System Properties.
In the Multimodal captioning section, select a Provider from the drop-down list.

A provider is the VLM you want to use for captioning.

Note: If your preferred provider is not listed, use the AI Control Tower (AICT) to configure approved third-party LLMs. For more information, see Configure third-party LLMs using AI Control Tower. NowLLM doesn't support this feature.

The corresponding Model options appear.
Select a Model from the drop-down list.

A model is a specific version of AI offered by the provider.
Select Save.

Result

Changes to the provider and model take effect immediately for multimodal captioning.

What to do next

AI Search administrators have additional configuration options available, as follows:

AI Search administrators can activate multimodal captioning for individual AI Search indexed sources. For details on this procedure, see Activate multimodal captioning for attachments from an indexed source.
Connector administrators can activate multimodal captioning in the crawl settings for an external content connector. To learn about configuring crawl settings for external content connectors, see Configuring crawl settings for external content connectors.

Parent Topic:Using AI Search Admin console

Keywords ServiceNow OttoAI Agentsgenerative AIagentic AI

64 lines · 3.8 KB Wrap lines

---
title: Configure multimodal captioning for AI Search
description: Use the AI Search Admin console to select the visual language model \(VLM\) provider and model for multimodal captioning.
locale: en-US
canonical_url: https://www.servicenow.com/docs/r/australia/platform-administration/search-administration/configure-multimodal-captioning-for-ai-search.html
release: australia
product: Search Administration
classification: search-administration
topic_type: task
last_updated: "2026-09-10"
reading_time_minutes: 2
keywords: [ServiceNow Otto, AI Agents, generative AI, agentic AI]
breadcrumb: [Using AI Search Admin console, AI Search Admin console, Search administration, Configure core features, Administer the ServiceNow AI Platform]
---

# Configure multimodal captioning for AI Search

Use the AI Search Admin console to select the visual language model \(VLM\) provider and model for multimodal captioning.

## Before you begin

- The Platform Multimodal Service plugin \(com.glide.platform\_mm\_service\) must be installed on your instance. If the plugin is not installed, multimodal captioning options will not appear in the AI Search Admin console. For more information, see [Activate the Platform Multimodal Service plugin](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/platform-administration/ai-search/activate-platform-multimodal-service-plugin.md).
- Role required: ais\_admin

## About this task

## Procedure

1. Navigate to **All** &gt; **AI Search Admin** &gt; **AI Search Admin Home**.

2. Select **System Properties**.

3. In the Multimodal captioning section, select a **Provider** from the drop-down list.

A provider is the VLM you want to use for captioning.

**Note:** If your preferred provider is not listed, use the AI Control Tower \(AICT\) to configure approved third-party LLMs. For more information, see [Configure third-party LLMs using AI Control Tower](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/intelligent-experiences/configure-third-party-llms-using-ai-control-tower.md). NowLLM doesn't support this feature.

The corresponding **Model** options appear.

4. Select a **Model** from the drop-down list.

A model is a specific version of AI offered by the provider.

5. Select **Save**.

## Result

Changes to the provider and model take effect immediately for multimodal captioning.

## What to do next

AI Search administrators have additional configuration options available, as follows:

- AI Search administrators can activate multimodal captioning for individual AI Search indexed sources. For details on this procedure, see [Activate multimodal captioning for attachments from an indexed source](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/platform-administration/ai-search/activate-multimodal-captioning.md).
- Connector administrators can activate multimodal captioning in the crawl settings for an external content connector. To learn about configuring crawl settings for external content connectors, see [Configuring crawl settings for external content connectors](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/platform-administration/search-administration/cfg-crawl-settings-ext-cont-connector.md).

**Parent Topic:**[Using AI Search Admin console](https://raw.githubusercontent.com/ServiceNow/ServiceNowDocs/australia/markdown/australia/platform-administration/search-administration/using-ais-admin-console.md)