Vertex AI Search offers two pricing models for custom search: a general consumption-based model and a configurable pricing model.
This page describes the pricing models and how to enable a configurable pricing model.
Pricing models
Vertex AI Search supports two pricing models:
- General: A pay-as-you-go consumption-based model.
- Configurable: A mixed-pricing model that lets you to unbundle capabilities and create a subscription package that is more closely aligned with your specific use cases. This pricing model lets you to pay only for the features that you need and benefit from their powerful capabilities. It offers two subscriptions and optional add-ons. After you enable this pricing model at project level, you can select configurable pricing for your apps and data stores.
Configurable pricing components
The configurable pricing model offers two subscriptions and different feature add-ons.
Subscription package
The subscription package has two subscriptions:
- Storage for your index, measured in GiB, for structured, unstructured, and website data.
- Search query for standard keyword search and filtering capabilities, measured in QPM (queries per minute).
Usage calculation
The storage size of the data stores for which configurable pricing applies is added up and compared against the storage subscription threshold. If the usage per hour is higher than the set threshold, the overage part is charged according to the general pricing model.
Similarly, the search query units (QPM) of the apps for which configurable pricing applies are added up and compared against the QPM subscription threshold. If the usage per minute is higher than the set threshold, the overage part is charged according to the general pricing model.
When the threshold is modified, the usage is compared to the new threshold for overage.
Increase subscription threshold
Within a month of enabling the configurable pricing, you can increase both storage size and QPM threshold after tracking the usage from the Project Billing Consumption tab.
Feature add-ons
You can enable powerful features by using add-ons.
Storage add-on
For the semantic embedding add-on, billing occurs when documents are imported to the data store.
Add-on |
Description |
|---|---|
| Semantic embedding | Enables the creation and maintenance of the vector embeddings for your documents, which is required before you can use the semantic query add-on in your search queries. |
Search request add-ons
Billing for these add-ons occurs when users invoke these features in a
search
API call.
Add-on |
Description |
|---|---|
| Semantic query | Enables semantic search capabilities for handling longer, more complex queries using embeddings. The semantic embedding add-on at the data store level is a prerequisite. |
| KPI & personalization | Enables user-event-based re-ranking and personalization to optimize for your business-specific KPIs. You are charged if the data quality has met the personalization tier and the training pipeline is ready. |
| AI overview | Enables generative features, such as generative summaries and follow-up questions. The semantic query add-on is a prerequisite. |
Enable configurable pricing for project
This section describes how to configure the pricing model at the project level. To use configurable pricing for an individual app or a data store, you must first enable it for the project.
Console
To enable configurable pricing using the Google Cloud console, follow these steps:
In the Google Cloud console, go to the Billing page.
On the Billing page, select Configurable Pricing.
Under Project-wide Configurable Pricing, input the QPM Subscription Threshold, the minimum is 1000 QPM/month.
Under Project-wide Configurable Pricing, input the Storage Subscription Threshold, the minimum is 50 GiB/month.
Click Save project settings.
To increase the thresholds, you can modify in the same page and click Save project settings.
Manage configurable pricing for an app
This section describes how to configure the pricing model at an app level. To use configurable pricing for an individual app, you must first enable it for the project. If you don't set up the configurable pricing for the app it uses the pay-as-you-go pricing model instead of being covered by the subscription cost.
Console
To enable configurable pricing for an app using the Google Cloud console, follow these steps:
In the Google Cloud console, go to the Apps page.
On the Apps page, click the app you want to modify.
In the main menu, click Billing page, and select Configurable Pricing.
Click Save.
Manage configurable pricing for data store
This section describes how to configure the pricing model at a data store level. To use configurable pricing for an individual data store, you must first enable it for the project. If you don't set up the configurable pricing for the data store, it uses the pay-as-you-go pricing model instead of being covered by the subscription cost.
Console
To enable configurable pricing for a data store using the Google Cloud console, follow these steps:
In the Google Cloud console, go to the Data stores page.
On the Data Stores page, click the data store you want to modify.
Click Billing page, and select Configurable Pricing.
Optional: Select Semantic (embedding) Add-on. This is a prerequisite for search serving for semantic query and AI Mode.
Click Save.
Manage configurable pricing for search query add-ons
Console
To select add-ons for an app using the Google Cloud console, follow these steps:
In the Google Cloud console, go to the AI Applications page.
On the Apps page, click the app that you want to modify.
Click Configurations.
On the UI tab, select the add-ons that you want. For apps of search type Search, the semantic query add-on can be disabled. The semantic query add-on is always on for Search with an answer and Search with follow-ups search types.
Click Save and publish.