You can manage Automated Embedding from the Atlas UI.
Enable or Disable Automated Embedding
To enable Automated Embedding, you must create a MongoDB Vector Search index with the autoEmbed type. To learn more, see Create an Automated
Embedding Index.
To disable Automated Embedding at the organization level, use an Atlas Resource Policy. You can disable Automated Embedding entirely or disable it with project-level exceptions.
To disable Automated Embedding entirely, use the following syntax in your Atlas Resource Policy:
forbid ( principal, action == ResourcePolicy::Action::"search.index.modify", resource ) when { context.search.index.isAutoEmbed };
To disable Automated Embedding with exceptions for specific projects, use the following syntax in your Atlas Resource Policy. Replace each <project-id> with the ID of the project to exempt. To find your project ID, see: Manage Project Settings.
forbid ( principal, action == ResourcePolicy::Action::"search.index.modify", resource ) when { context.search.index.isAutoEmbed } unless { resource in ResourcePolicy::Project::"<project-id-1>" || resource in ResourcePolicy::Project::"<project-id-2>" };
When active, these policies prevent new autoEmbed type indexes from being created. Indexes created before the resource policy was activated are not affected and must be deleted manually to maintain compliance.
To apply this policy to your organization:
To enable Automated Embedding, you must do the following:
Create a Voyage AI API key for indexing operations.
To learn more, see Model API Keys.
Create a Voyage AI API key for query operations.
To learn more, see Model API Keys.
Configure mongot to use the API keys.
To learn more, see Voyage AI API Keys.
Create an autoEmbed type MongoDB Vector Search index.
To learn more, see Create an Automated Embedding Index.
View Automated Embedding Usage
To view aggregated token usage and embedding model requests for Automated Embedding service, you can use MongoDB Vector Search dashboard or Atlas metrics.
Monitor Dashboard
Log in to Atlas.
In Atlas, go to the Search & Vector Search page for your cluster.
You can go the MongoDB Search page from the Search & Vector Search option, or the Data Explorer.
If it's not already displayed, select the organization that contains your project from the Organizations menu in the navigation bar.
If it's not already displayed, select your project from the Projects menu in the navigation bar.
In the sidebar, click Search & Vector Search under the Database heading.
If your project has multiple clusters, select the cluster you want to use from the Select cluster dropdown, then click Go to Atlas Search.
The Search & Vector Search page displays.
If it's not already displayed, select the organization that contains your project from the Organizations menu in the navigation bar.
If it's not already displayed, select your project from the Projects menu in the navigation bar.
In the sidebar, click Data Explorer under the Database heading.
Expand the database and select the collection.
Click the Indexes tab for the collection.
Click the Search and Vector Search link in the banner.
The Search & Vector Search page displays.
Click Usage under Automated Embedding.
The Automated Embedding Usage page displays. The Automated Embedding Usage page displays the following information:
Section | Description |
|---|---|
Usage Summary | Overview of your Automated Embedding usage, including total tokens used and remaining free tokens. |
Usage by Model or Index | Breakdown of total token usage by embedding model or selected index during a selected time period. |
Usage by Operation | Breakdown of token requests by operation type (indexing and querying). |
The following procedure only applies to API keys created from the Atlas UI. For API keys created directly from Voyage AI, see Rate Limits.
Log in to Atlas.
View the token usage and embedding model requests for Automated Embedding.
The Usage page displays usage metrics for all model API keys in your organization. The page provides the following tabs:
Usage Activity: Shows detailed usage metrics and charts for your organization.
Free Usage: Shows free quota consumption by model.
Paid Usage: Shows paid usage by model and API key.
To learn more, see Usage Page Features.
Monitor Metrics
To monitor Automated Embedding usage using metrics:
Log in to Atlas.
Go to the Metrics view for your cluster.
If it's not already displayed, select your desired organization from the Organizations menu in the navigation bar.
If it's not already displayed, select your desired project from the Projects menu in the navigation bar.
Click on your cluster name to open the Cluster sidebar.
In the sidebar, click Metrics.
Select the Auto Embedding Metrics to review.
You can monitor Automated Embedding usage through the following metrics in the Auto Embedding Metrics section:
Metric | Description |
|---|---|
Auto Embedding Index Embedding Request | The Auto Embedding Index Embedding Request graph displays the number of requests to generate embeddings for indexing operations. Rates are shown for each embedding model in use. |
Auto Embedding Index Tokens | The Auto Embedding Index Tokens graph displays the number of tokens used for indexing operations. |
Auto Embedding Query Embedding Request | The Auto Embedding Query Embedding Request graph displays the number of requests to generate embeddings for query operations. |
Auto Embedding Query Tokens | The Auto Embedding Query Tokens graph displays the number of tokens used for query operations. |
Auto Embedding Replication Lag | The Auto Embedding Replication Lag graph displays the replication lag in milliseconds for Automated Embedding indexing operations. |
For each cluster, Atlas displays a single graph that aggregates token usage across all Automated Embedding indexes on the cluster. The graph includes token usage from both the initial sync and steady-state updates, and displays a separate line for each embedding model that your indexes use so you can compare token usage by model.
If you deployed Search Nodes for workload isolation, Atlas displays this graph in the Search Nodes metrics view for each Search Node instead of the cluster metrics view.
Automated Embedding metrics are not available for self-managed deployments.
Manage Rate Limits
You can manage rate limits for Automated Embedding from the Atlas UI.
View Rate Limits
To view rate limits for Automated Embedding:
Log in to Atlas.
Go to the Search & Vector Search page for your project.
If it's not already displayed, select your desired organization from the Organizations menu in the navigation bar.
If it's not already displayed, select your desired project from the Projects menu in the navigation bar.
In the sidebar, click Search & Vector Search under the Database header.
The Search & Vector Search page displays.
The following procedure only applies to API keys created from the Atlas UI. For API keys created directly from Voyage AI, see Rate Limits.
Log in to Atlas.
Increase Rate Limits
If you are on the Free Tier and need to increase rate limits for Automated Embedding, add a payment method to upgrade to a paid tier. To learn more, see Set Payment Method.
If you are on a Paid Tier, MongoDB automatically increases rate limits as your usage increases. If you still need a higher rate limit, contact MongoDB Support.