AI Agent Engineering > Create connections and tools > Embedding model connections

Embedding model connections

An embedding model connection is a connection to a model provider that allows your AI agent to retrieve text data for advanced searches and Retrieval-Augmented Generation (RAG) applications. These connections are used in some tool connections such as Pinecone and Azure AI Search.

Create an embedding model connection for each embedding model that your AI agents will use. The type of connection you configure is based on the model provider. For example, to create a connection to Amazon Titan Embeddings G1 - Text, configure an Amazon Bedrock embedding model connection.

The following table lists the supported embedding models and model providers:

Model	Model provider
Amazon Titan Embeddings G1 - Text	Amazon Bedrock
Amazon Titan Text Embeddings V2	Amazon Bedrock
AzureAI text-embedding-3-large	Azure Open AI
AzureAI text-embedding-3-large	Azure Open AI
AzureAI text-embedding-ada-002	Azure Open AI
Cohere Embed - English	Amazon Bedrock
Google Gemini embedding-001	Google Gemini
Google Gemini gemini-embedding-001	Google Gemini
Google Gemini text-embedding-004	Google Gemini

AzureAI embedding model connection details

You need to configure a AzureAI embedding model connection before you can use it in an AI agent.

The following table describes the properties of an AzureAI connection. These fields apply to every version of AzureAI:

Property	Description
Name	Name of the AzureAI embedding model connection.
Location	Project or folder to save your assets. By default, assets are saved to the Default project.
Embedding Model Provider	Provider of the embedding model being configured.
Description	Optional. Description of the connection to the model.
Embedding Model API Key	Authentication credential to securely access the Azure AI embedding model.
Deployment Name	Identifier or name of the specific embedded AI model or deployment you're connecting to.
Azure Endpoint	Azure endpoint URL specific to your Azure AI service instance.

Amazon Bedrock Embedding model connection details

You need to configure an Amazon Bedrock embedding model connection before you can use it in an AI agent.

The following table describes the properties of an Amazon Bedrock embedding model connection. These properties apply to every type of Bedrock embedding model:

Property	Description
Name	Name of the Amazon Bedrock embedding model connection.
Location	Project or folder to save your assets. By default, assets are saved to the Default project.
Embedding Model Provider	Provider of the embedding model being configured.
Description	Optional. Description of the embedding model connection.
Embedding Model	ID string of the embedding model you want your AI agents to use for converting text into vector embeddings. If you use a cross-region inference type embedding model, add the region prefix at the beginning of the model ID string. For example, to connect to Cohere Embed-English in the United States, enter the following model ID: us.cohere.embed-v4:0
Region	AWS region where your Amazon Bedrock service is deployed, for example: us-east-1
Access Key	Access key ID for the IAM user that has permissions to access Bedrock embedding services.
Secret Key	Secret access key for the IAM user that has permissions to access Bedrock embedding services.

Google Gemini embedding model connection details

You need to configure a Google Gemini embedding model connection before you can use it in an AI agent.

The following table describes the properties of a Google Gemini connection. These fields apply to every version of Google Gemini:

Property	Description
Name	Name of the Google Gemini embedding model connection.
Location	Project or folder to save your assets. By default, assets are saved to the Default project.
Embedding Model Provider	Provider of the embedding model being configured.
Description	Optional. Description of the connection to the model.
Embedding Model API Key	Credential that authenticates your connection and authorizes your AI agent to access Google Gemini’s embedding model services. You obtain this key from the Google Cloud Console.
Embedding Model	Name of the embedding model to use.

Creating an embedding model connection

Create at least one embedding model connection before you configure an AI agent.

1From the navigation menu, click New.

2In the New Asset dialog box, select Embedding Model from the New Asset list.

3Select an embedding model provider and then click Create.

4Enter the connection details.

5Click Save.