Vertex AI pricing

Prices are listed in US Dollars (USD). If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

Vertex AI pricing compared to legacy product pricing

The costs for Vertex AI remain the same as they are for the legacy AI Platform and AutoML products that Vertex AI supersedes, with the following exceptions:

• Legacy AI Platform Prediction and AutoML Tables predictions supported lower-cost, lower-performance machine types that aren't supported for Vertex AI Prediction and AutoML tabular.

• Legacy AI Platform Prediction supported scale-to-zero, which isn't supported for Vertex AI Prediction.

Vertex AI also offers more ways to optimize costs, such as the following:

• Optimized TensorFlow runtime.

• Support for co-hosting models.

• No minimum usage duration for Training and Prediction. Instead, usage is charged in 30 second increments.

Pricing for Generative AI on Vertex AI

For Generative AI on Vertex AI pricing information, see Pricing for Generative AI on Vertex AI.

Pricing for AutoML models

For Vertex AI AutoML models, you pay for three main activities:

• Training the model
• Deploying the model to an endpoint
• Using the model to make predictions

Vertex AI uses predefined machine configurations for Vertex AutoML models, and the hourly rate for these activities reflects the resource usage.

The time required to train your model depends on the size and complexity of your training data. Models must be deployed before they can provide online predictions or online explanations.

You pay for each model deployed to an endpoint, even if no prediction is made. You must undeploy your model to stop incurring further charges. Models that are not deployed or have failed to deploy are not charged.

You pay only for compute hours used; if training fails for any reason other than a user-initiated cancellation, you are not billed for the time. You are charged for training time if you cancel the operation.

Select a model type below for pricing information.

Prediction charges for Vertex Explainable AI

Compute associated with Vertex Explainable AI is charged at same rate as prediction. However, explanations take longer to process than normal predictions, so heavy usage of Vertex Explainable AI along with auto-scaling could result in more nodes being started, which would increase prediction charges.

Vertex AI Forecast

Custom-trained models

Training

The tables below provide the approximate price per hour of various training configurations. You can choose a custom configuration of selected machine types. To calculate pricing, sum the costs of the virtual machines you use.

If you use Compute Engine machine types and attach accelerators, the cost of the accelerators is separate. To calculate this cost, multiply the prices in the table of accelerators below by how many machine hours of each type of accelerator you use.

Machine types

^*This amount includes GPU price, since this instance type always requires a fixed number of GPU accelerators.
If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

Accelerators

If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

* The price for training using a Cloud TPU Pod is based on the number of cores in the Pod. The number of cores in a pod is always a multiple of 32. To determine the price of training on a Pod that has more than 32 cores, take the price for a 32-core Pod, and multiply it by the number of cores, divided by 32. For example, for a 128-core Pod, the price is (32-core Pod price) * (128/32). For information about which Cloud TPU Pods are available for a specific region, see System Architecture in the Cloud TPU documentation.

Disks

If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

• All use is subject to the Vertex AI quota policy.
• You are required to store your data and program files in Google Cloud Storage buckets during the Vertex AI lifecycle. See more about Cloud Storage usage.

You are charged for training your models from the moment when resources are provisioned for a job until the job finishes.

Scale tiers for predefined configurations (AI Platform Training)

You can control the type of processing cluster to use when training your model. The simplest way is to choose from one of the predefined configurations called scale tiers. Read more about scale tiers.

Machine types for custom configurations

If you use Vertex AI or select CUSTOM as your scale tier for AI Platform Training, you have control over the number and type of virtual machines to use for the cluster's master, worker and parameter servers. Read more about machine types for Vertex AI and machine types for AI Platform Training.

The cost of training with a custom processing cluster is the sum of all the machines you specify. You are charged for the total time of the job, not for the active processing time of individual machines.

Gen AI Evaluation Service

Vertex AI Gen AI Evaluation Service charges string input and output fields by every 1,000 characters. One character is defined as one Unicode character. White space is excluded from the count. Failed evaluation request, including filtered response, will not be charged for input nor output. At the end of each billing cycle, fractions of one cent ($0.01) are rounded to one cent.

Gen AI Evaluation Service is generally available (GA). Pricing took effect on September 27, 2024.

Metric	Pricing
Pointwise	Input: $0.005 per 1k characters Output: $0.015 per 1k characters
Pairwise	Input: $0.005 per 1k characters Output: $0.015 per 1k characters

Computation-based metrics are charged at $0.00003 per 1k characters for input and $0.00009 per 1k characters for output. They are referred to as Automatic Metric in SKU.

Metric Name	Type
Exact Match	Computation-based
Bleu	Computation-based
Rouge	Computation-based
Tool Call Valid	Computation-based
Tool Name Match	Computation-based
Tool Parameter Key Match	Computation-based
Tool Parameter KV Match	Computation-based

Prices are listed in US Dollars (USD). If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

Legacy model-based metrics are charged at $0.005 per 1k characters for input and $0.015 per 1k characters for output.

Metric Name	Type
Coherence	Pointwise
Fluency	Pointwise
Fulfillment	Pointwise
Safety	Pointwise
Groundedness	Pointwise
Summarization Quality	Pointwise
Summarization Helpfulness	Pointwise
Summarization Verbosity	Pointwise
Question Answering Quality	Pointwise
Question Answering Relevance	Pointwise
Question Answering Helpfulness	Pointwise
Question Answering Correctness	Pointwise
Pairwise Summarization Quality	Pairwise
Pairwise Question Answering Quality	Pairwise

Prices are listed in US Dollars (USD). If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

Ray on Vertex AI

Training

The tables below provide the approximate price per hour of various training configurations. You can choose a custom configuration of selected machine types. To calculate pricing, sum the costs of the virtual machines you use.

If you use Compute Engine machine types and attach accelerators, the cost of the accelerators is separate. To calculate this cost, multiply the prices in the table of accelerators below by how many machine hours of each type of accelerator you use.

Machine types

If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

Accelerators

If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

* The price for training using a Cloud TPU Pod is based on the number of cores in the Pod. The number of cores in a pod is always a multiple of 32. To determine the price of training on a Pod that has more than 32 cores, take the price for a 32-core Pod, and multiply it by the number of cores, divided by 32. For example, for a 128-core Pod, the price is (32-core Pod price) * (128/32). For information about which Cloud TPU Pods are available for a specific region, see System Architecture in the Cloud TPU documentation.

Disks

If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

• All use is subject to the Vertex AI quota policy.
• You are required to store your data and program files in Google Cloud Storage buckets during the Vertex AI lifecycle. See more about Cloud Storage usage.

You are charged for training your models from the moment when resources are provisioned for a job until the job finishes.

Prediction and explanation

The following tables provide the prices of batch prediction, online prediction, and online explanation per node hour. A node hour represents the time a virtual machine spends running your prediction job or waiting in an active state (an endpoint with one or more models deployed) to handle prediction or explanation requests.

Choose a region to view its pricing table:

• The Americas
• Europe
• Asia Pacific
• Middle east

Pricing for the Americas

The following tables provide the price per node hour for each machine type.

TPU v5e
ct5lp-hightpu-1t	Approximations:
us-west1	$1.38
ct5lp-hightpu-4t	Approximations:
us-west1	$5.52
ct5lp-hightpu-8t	Approximations:
us-west1	$11.04

Pricing for Europe

The following tables provide the price per node hour for each machine type.

Pricing for Asia Pacific

The following tables provide the price per node hour for each machine type.

Pricing for the Middle East

Each machine type is charged as the following SKUs on your Google Cloud bill:

• vCPU cost: measured in vCPU hours
• RAM cost: measured in GB hours
• GPU cost: if either built into the machine or optionally configured, measured in GPU hours

The prices for machine types are used to approximate the total hourly cost for each prediction node of a model version using that machine type.

For example, a machine type of n1-highcpu-32 includes 32 vCPUs and 32 GB of RAM. Therefore, the hourly pricing equals 32 vCPU hours + 32 GB hours.

The SKU pricing table is available by region. Each table shows the vCPU, RAM, and built-in GPU pricing for prediction machine types, which more precisely reflect the SKUs charged.

To view the SKU pricing per region, choose a region to view its pricing table:

• The Americas
• Europe
• Asia Pacific
• Middle East

SKU pricing for the Americas

SKU pricing for Europe

SKU pricing for Asia Pacific

SKU pricing for Middle East

Some machine types allow you to add optional GPU accelerators for prediction. Optional GPUs incur an additional charge, separate from those described in the previous table. View each pricing table, which describes the pricing for each type of optional GPU.

Pricing is per GPU. If you use multiple GPUs per prediction node (or if your version scales to use multiple nodes),the costs scale accordingly.

AI Platform Prediction serves predictions from your model by running a number of virtual machines ("nodes"). By default, Vertex AI automatically scales the number of nodes running at any time. For online prediction, the number of nodes scales to meet demand. Each node can respond to multiple prediction requests. For batch prediction, the number of nodes scales to reduce the total time it takes to run a job. You can customize how prediction nodes scale.

You are charged for the time that each node runs for your model, including:

• When the node is processing a batch prediction job.
• When the node is processing an online prediction request.
• When the node is in a ready state for serving online predictions.

The cost of one node running for one hour is a node hour. The table of prediction prices describes the price of a node hour, which varies across regions and between online prediction and batch prediction.

You can consume node hours in fractional increments. For example, one node running for 30 minutes costs 0.5 node hours.

Cost calculations for Compute Engine (N1) machine types

• The running time of a node is billed in 30-second increments. This means that every 30 seconds, your project is billed for 30 seconds worth of whatever vCPU, RAM, and GPU resources that your node is using at that moment.

More about automatic scaling of prediction nodes

Online prediction	Batch prediction
The priority of the scaling is to reduce the latency of individual requests. The service keeps your model in a ready state for a few idle minutes after servicing a request.	The priority of the scaling is to reduce the total elapsed time of the job.
Scaling affects your total charges each month: the more numerous and frequent your requests, the more nodes will be used.	Scaling should have little effect on the price of your job, though there is some overhead involved in bringing up a new node.
You can choose to let the service scale in response to traffic (automatic scaling) or you can specify a number of nodes to run constantly to avoid latency (manual scaling). If you choose automatic scaling, the number of nodes scales automatically. For AI Platform Prediction legacy (MLS1) machine type deployments, the number of nodes can scale down to zero for no-traffic durations. Vertex AI deployments and other types of AI Platform Prediction deployments cannot scale down to zero nodes. If you choose manual scaling, you specify a number of nodes to keep running all the time. You are charged for all of the time that these nodes are running, starting at the time of deployment and persisting until you delete the model version.	You can affect scaling by setting a maximum number of nodes to use for a batch prediction job, and by setting the number of nodes to keep running for a model when you deploy it.

Batch prediction jobs are charged after job completion

Batch prediction jobs are charged after job completion, not incrementally during the job. Any Cloud Billing budget alerts that you have configured aren't triggered while a job is running. Before starting a large job, consider running some cost benchmark jobs with small input data first.

Example of a prediction calculation

A real-estate company in an Americas region runs a weekly prediction of housing values in areas it serves. In one month, it runs predictions for four weeks in batches of 3920, 4277, 3849, and 3961. Jobs are limited to one node and each instance takes an average of 0.72 seconds of processing.

First calculate the length of time that each job ran:

3920 instances * (0.72 seconds / 1 instance) * (1 minute / 60 seconds) = 47.04 minutes
4277 instances * (0.72 seconds / 1 instance) * (1 minute / 60 seconds) = 51.324 minutes
3849 instances * (0.72 seconds / 1 instance) * (1 minute / 60 seconds) = 46.188 minutes
3961 instances * (0.72 seconds / 1 instance) * (1 minute / 60 seconds) = 47.532 minutes

Each job ran for more than ten minutes, so it is charged for each minute of processing:

($0.0909886 / 1 node hour) * (1 hour / 60 minutes) * 48 minutes * 1 node = $0.0632964
($0.0909886 / 1 node hour) * (1 hour / 60 minutes) * 52 minutes * 1 node = $0.0685711
($0.0909886 / 1 node hour) * (1 hour / 60 minutes) * 47 minutes * 1 node = $0.061977725
($0.0909886 / 1 node hour) * (1 hour / 60 minutes) * 48 minutes * 1 node = $0.0632964

The total charge for the month is $0.26.

This example assumed jobs ran on a single node and took a consistent amount of time per input instance. In real usage, make sure to account for multiple nodes and use the actual amount of time each node spends running for your calculations.

Charges for Vertex Explainable AI

Feature-based explanations

Feature-based explanations come at no extra charge to prediction prices. However, explanations take longer to process than normal predictions, so heavy usage of Vertex Explainable AI along with auto-scaling could result in more nodes being started, which would increase prediction charges.

Example-based explanations

Pricing for example-based explanations consists of the following:

• When you upload a model or update a model's dataset, you are billed:

  • per node hour for the batch prediction job that is used to generate the latent space representations of examples. This is billed at the same rate as prediction.
  • a cost for building or updating indexes. This cost is the same as the indexing costs for Vector Search, which is number of examples * number of dimensions * 4 bytes per float * $3.00 per GB. For example, if you have 1 million examples and 1,000 dimension latent space, the cost is $12 (1,000,000 * 1,000 * 4 * 3.00 / 1,000,000,000).

• When you deploy to an endpoint, you are billed per node hour for each node in your endpoint. All compute associated with the endpoint is charged at same rate as prediction. However, because Example-based explanations require additional compute resources to serve the Vector Search index, this results in more nodes being started which increases prediction charges.

Vertex AI Neural Architecture Search

The following tables summarize the pricing in each region where Neural Architecture Search is available.

Prices

The following tables provide the price per hour of various configurations.

You can choose a predefined scale tier or a custom configuration of selected machine types. If you choose a custom configuration, sum the costs of the virtual machines you use.

Accelerator-enabled legacy machine types include the cost of the accelerators in their pricing. If you use Compute Engine machine types and attach accelerators, the cost of the accelerators is separate. To calculate this cost, multiply the prices in the following table of accelerators by the number of each type of accelerator you use.

Machine types

If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

Prices for a2-highgpu instances include the charges for the attached NVIDIA_TESLA_A100 Accelerators.

Accelerators

If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

Disks

If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

Notes:

1. All use is subject to the Neural Architecture Search quota policy.
2. You are required to store your data and program files in Cloud Storage buckets during the Neural Architecture Search lifecycle. See more about Cloud Storage usage.
3. For volume-based discounts, contact the Sales team.
4. The disk price is only charged when you configure the disk size of each VM to be larger than 100 GB. There is no charge for the first 100 GB (the default disk size) of disk for each VM. For example, if you configure each VM to have 105 GB of disk, then you are charged for 5 GB of disk for each VM.

Required use of Cloud Storage

In addition to the costs described in this document, you are required to store data and program files in Cloud Storage buckets during the Neural Architecture Search lifecycle. This storage is subject to the Cloud Storage pricing policy.

Required use of Cloud Storage includes:

• Staging your training application package.

• Storing your training input data.

• Storing the output of your jobs. Neural Architecture Search doesn't require long-term storage of these items. You can remove the files as soon as the operation is complete.

Free operations for managing your resources

The resource management operations provided by Neural Architecture Search are available free of charge. The Neural Architecture Search quota policy does limit some of these operations.

Resource	Free operations
jobs	get, list, cancel
operations	get, list, cancel, delete

Vertex AI Pipelines

Vertex AI Pipelines charges a run execution fee of $0.03 per Pipeline Run. You are not charged the execution fee during the Preview release. You also pay for Google Cloud resources you use with Vertex AI Pipelines, such as Compute Engine resources consumed by pipeline components (charged at the same rate as for Vertex AI training). Finally, you are responsible for the cost of any services (such as Dataflow) called by your pipeline.

Vertex AI Feature Store

Vertex AI Feature Store is Generally Available (GA) since November 2023. For information on the previous version of the product go to Vertex AI Feature Store (Legacy).

New Vertex AI Feature Store

The new Vertex AI Feature Store supports functionality across 2 types of operations:

• Offline operations are operations to transfer, store, retrieve and transform data in the offline store (BigQuery)
• Online operations are operations to transfer data into the online store(s) and operations on data while it is in the online store(s).

Offline Operations Pricing

Since BigQuery is used for offline operations, please refer to BigQuery pricing for functionality such as ingestion to the offline store, querying the offline store, and offline storage.

Online Operations Pricing

For online operations, Vertex AI Feature Store charges for any GA features to transfer data into the online store, serve data or store data. A node-hour represents the time a virtual machine spends to complete an operation, charged to the minute.

Optimized online serving and Bigtable online serving use different architectures, so their nodes are not comparable. If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

Online Operations Workload Estimates

Consider the following guidelines when estimating your workloads. The number of nodes required for a given workload may differ across each serving approach.

• Data processing:
  • Ingestion - One node can ingest approximately a minimum of 100 MB of data per hour into a Bigtable Online Store or an Optimized Online Store if no analytical functions are used.
• Bigtable online serving: Each node can support approximately 15,000 QPS and up to 5TB of storage.
• Optimized online serving: Performance is based on the machine type and replicas, which are automatically configured to minimize costs subject to the workload. Each node can have a minimum of 2 and a maximum of 6 replicas for high availability and autoscaling. You're charged for the number of replicas accordingly. For more details, see the example monthly scenarios.
  • For non embeddings-related workloads, each node can support approximately 500 QPS and up to 200GB of storage.
  • For embeddings-related workloads, each node can support approximately 500 QPS and up to 4 GB of storage of 512 dimensional data.

You can view the number of nodes (with replicas) in the Metric Explorer:

Example Monthly Scenarios (assuming us-central1)

Data streaming workload - Bigtable online serving with 2.5TB of data (1GB refreshed daily) and 1200QPS

Operations	Monthly Usage	Monthly Cost
Data processing node	(1 GB/day) * (30 days/month) * (1,000 MB/GB) * (1 node-hr / 100MB) = 300 node-hr	300 node-hr * ($0.08 per node-hr) = $24
Optimized online serving node	N/A	N/A
Bigtable online serving node	(1 node) * (24 hr/day) * (30 days/month) = 720 node-hr	720 node-hr * ($0.94 per node-hr) = $677
Bigtable online serving storage	(2.5 TB-month) * (1000 GB/TB) = 2500 GB-month	2500 GB-month * ($0.25 per GB-month) = $625
Total		$1,326

High QPS workload - Optimized online serving with 10GB of non-embedding data (5GB refreshed daily) and 2000QPS

Operations	Monthly Usage	Monthly Cost
Data processing node	(5 GB/day) * (30 days/month) * (1,000 MB/GB) * (1 node-hr / 100MB) = 1500 node-hr	1500 node-hr * ($0.08 per node-hr) = $120
Optimized online serving node	Roundup(10GB * (1 node / 200 GB)) = 1 * max(2 default replicas, 2000 QPS * (1 replica / 500 QPS)) = 4 total nodes * (24 hr/day) * (30days/month) =2880 node-hr	2880 node-hr * (0.30 per node-hr) = $864
Bigtable online serving node	N/A	N/A
Bigtable online serving storage	N/A	N/A
Total		$984

Embeddings serving workload - Optimized online serving with 20GB of embeddings data (2GB refreshed daily) and 800QPS

Operations	Monthly Usage	Monthly Cost
Data processing node	(2 GB/day) * (30 days/month) * (1,000 MB/GB) * (1 node-hr / 100MB) = 600 node-hr	600 node-hr * ($0.08 per node-hr) = $48
Optimized online serving node	Roundup(20GB* (1 node / 4GB) = 5 * max(2 default replicas, 800 QPS * (1 replica / 500 QPS)) = 10 total nodes * (24 hr/day) * (30days/month) = 7200 node-hr	7200 node-hr * (0.30 per node-hr) = $2160
Bigtable online serving node	N/A	N/A
Bigtable online serving storage	N/A	N/A
Total		$2,208

Vertex AI Feature Store (Legacy)

Prices for Vertex AI Feature Store (Legacy) are based on the amount of feature data in online and offline storage as well as the availability of online serving. A node per hour represents the time a virtual machine spends serving feature data or waiting in a ready state to handle feature data requests.

If you pay in a currency other than USD, the prices listed in your currency on Cloud Platform SKUs apply.

When you enable feature value monitoring, billing includes applicable charges above in addition to applicable charges that follow:

• $3.50 per GB for all data analyzed. With snapshot analysis enabled, snapshots taken for data in Vertex AI Feature Store (Legacy) are included. With import feature analysis enabled, batches of ingested data are included.
• Additional charges for other Vertex AI Feature Store (Legacy) operations used with feature value monitoring include the following:
  • The snapshot analysis feature periodically takes a snapshot of the feature values based on your configuration for the monitoring interval.
  • The charge for a snapshot export is the same as a regular batch export operation.

Snapshot Analysis Example

A data scientist enables feature value monitoring for their Vertex AI Feature Store (Legacy) and turns on monitoring for a daily snapshot analysis. A pipeline runs daily for the entity types monitoring. The pipeline scans 2GB of data in Vertex AI Feature Store (Legacy) and exports a snapshot containing 0.1GB of data. The total charge for one day's analysis is:

(0.1 GB * $3.50) + (2 GB * $0.005) = $0.36

Ingestion Analysis Example

A data scientist enables feature value monitoring for their Vertex AI Feature Store (Legacy) and turns on monitoring for ingestion operations. An ingestion operation imports 1GB of data into Vertex AI Feature Store (Legacy). The total charge for feature value monitoring is:

(1 GB * $3.50) = $3.50

Vertex ML Metadata

Metadata storage is measured in binary gigabytes (GiB), where 1 GiB is 1,073,741,824 bytes. This unit of measurement is also known as a gibibyte.

Vertex ML Metadata charges $10 per gibibyte (GiB) per month for metadata storage. Prices are pro-rated per megabyte (MB). For example, if you store 10 MB of metadata, you are charged $0.10 per month for that 10 MB of metadata.

Prices are the same in all regions where Vertex ML Metadata is supported.

Vertex AI TensorBoard

To use Vertex AI TensorBoard, request that the IAM administrator of the project assign you to the role "Vertex AI TensorBoard Web App User". The Vertex AI Administrator role also has access.

Beginning in August 2023, Vertex AI TensorBoard pricing changed from a per-user monthly license of $300/month to $10 GiB/month for data storage of logs and metrics. This means no more subscription fees. You will only pay for the storage you’ve used. See the Vertex AI TensorBoard: Delete Outdated TensorBoard Experiments tutorial for how to manage storage.

Vertex AI Vizier

Vertex AI Vizier is a black-box optimization service inside Vertex AI. The Vertex AI Vizier pricing model consists of the following:

• There is no charge for trials that use RANDOM_SEARCH and GRID_SEARCH. Learn more about the search algorithms.
• The first 100 Vertex AI Vizier trials per calendar month are available at no charge (trials using RANDOM_SEARCH and GRID_SEARCH do not count against this total).
• After 100 Vertex AI Vizier trials, subsequent trials during the same calendar month are charged at $1 per trial (trials that use RANDOM_SEARCH or GRID_SEARCH incur no charges).

Vector Search

Pricing for Vector Search Approximate Nearest Neighbor service consists of:

• Per node hour pricing for each VM used to host a deployed index.
• A cost for building new indexes, updating existing indexes, and using streaming index updates.

Data processed during building and updating indexes is measured in binary gigabytes (GiB), where 1 GiB is 1,073,741,824 bytes. This unit of measurement is also known as a gibibyte.

Vector Search charges $3.00 per gibibyte (GiB) of data processed in all regions. Vector Search charges $0.45/GiB ingested for Streaming Update inserts.

The following tables summarize the pricing of an index serving in each region where the Vector Search is available. The price corresponds to the machine type, by region, and is charged per node hour.

Vector Search pricing examples

Vector Search pricing is determined by the size of your data, the amount of queries per second (QPS) you want to run, and the number of nodes you use. To get your estimated serving cost, you need to calculate your total data size. Your data size is the number of your embeddings/vectors* the number of dimensions you have* 4 bytes per dimension. After you have the size of your data you can calculate the serving cost and the building cost. The serving cost plus the building cost equals your monthly total cost.

• Serving cost: # replicas/shard * # shards (~data size/shard size) * hourly cost * 730 hours
• Building cost: data size(in GiB) * $3/GiB * # of updates/month

Streaming update: Vector Search uses heuristics-based metrics to determine when to trigger compaction. If the oldest uncompacted data is five days old, compaction is always triggered. You are billed for the cost of rebuilding the index at the same rate of a batch update, in addition to the streaming update costs.

Number of embeddings/vectors	Number of dimensions	Queries per second (QPS)	Machine Type	Nodes	Estimated monthly serving cost
2 million	128	100	e2-standard-2	1	$68
20 million	256	1,000	e2-standard-16	1	$547
20 million	256	3,000	e2-standard-16	3	$1,642
100 million	256	500	e2-highmem-16	2	$1,477
1 billion	100	500	e2-highmem-16	8	$5,910

All examples are based on machine types in us-central1. The cost you incur will vary with recall rate and latency requirements. The estimated monthly serving cost is directly related to the number of nodes used in the console. To learn more about configuration parameters that affect cost, see Configuration parameters which affect recall and latency.

If you have high queries per second (QPS), batching these queries can reduce total costs up to 30%-40%.

Vertex AI Model Registry

The Vertex AI Model Registry is a central repository which tracks and lists your models and model versions. You can import models into Vertex AI and they appear in the Vertex AI Model Registry. There is no cost associated with having your models in the Model Registry. Cost is only incurred when you deploy the model to an endpoint or perform a batch prediction on the model. This cost is determined by the type of model you are deploying.

To learn more about pricing for deploying custom models from the Vertex AI Model Registry, see Custom-trained models. To learn more about pricing for deploying AutoML models, see Pricing for AutoML models.

Vertex AI Model Monitoring

Vertex AI enables you to monitor the continued effectiveness of your model after you deploy it to production. For more information, see Introduction to Vertex AI Model Monitoring.

When you use Vertex AI Model Monitoring, you are billed for the following:

• $3.50 per GB for all data analyzed, including the training data provided and prediction data logged in a BigQuery table.
• Charges for other Google Cloud products that you use with Model Monitoring, such as BigQuery storage or Batch Explain when attribution monitoring is enabled.

Vertex AI Model Monitoring is supported in the following regions: us-central1, europe-west4, asia-east1, and asia-southeast1. Prices are the same for all regions.

Data sizes are measured after they are converted to TfRecord format.

Training datasets incur a one-time charge when you set up a Vertex AI Model Monitoring job.

Prediction Datasets consist of logs collected from the Online Prediction service. As prediction requests arrive during different time windows, the data for each time window is collected and the sum of the data analyzed for each prediction window is used to calculate the charge.

Example: A data scientist runs model monitoring on the prediction traffic belonging to their model.

• The model is trained from a BigQuery dataset. The data size after converting to TfRecord is 1.5GB.
• Prediction data logged between 1:00 - 2:00 p.m. is 0.1 GB, between 3:00 - 4:00 p.m. is 0.2 GB.

• The total price for setting up the model monitoring job is:

  (1.5 GB * $3.50) + ((0.1 GB + 0.2 GB) * $3.50) = $6.30

Vertex AI Workbench

Select instances, managed notebooks, or user-managed notebooks for pricing information.

Additional Google Cloud resources

In addition to the costs mentioned previously, you also pay for any Google Cloud resources that you use. For example:

• Data analysis services: You incur BigQuery costs when you issue SQL queries within a notebook (see BigQuery pricing).

• Customer-managed encryption keys: You incur costs when you use customer-managed encryption keys. Each time your managed notebooks or user-managed notebooks instance uses a Cloud Key Management Service key, that operation is billed at the rate of Cloud KMS key operations (see Cloud Key Management Service pricing).

Colab Enterprise

For Colab Enterprise pricing information, see Colab Enterprise pricing.

Deep Learning Containers, Deep Learning VM, and AI Platform Pipelines

For Deep Learning Containers, Deep Learning VM Images, and AI Platform Pipelines, pricing is calculated based on the compute and storage resources that you use. These resources are charged at the same rate you currently pay for Compute Engine and Cloud Storage.

In addition to the compute and storage costs, you also pay for any Google Cloud resources that you use. For example:

• Data analysis services: You incur BigQuery costs when you issue SQL queries within a notebook (see BigQuery pricing).

• Customer-managed encryption keys: You incur costs when you use customer-managed encryption keys. Each time your managed notebooks or user-managed notebooks instance uses a Cloud Key Management Service key, that operation is billed at the rate of Cloud KMS key operations (see Cloud Key Management Service pricing).

Data labeling

Vertex AI enables you to request human labeling for a collection of data that you plan to use to train a custom machine learning model. Prices for the service are computed based on the type of labeling task.

• For regular labeling tasks, the prices are determined by the number of annotation units.
  • For an image classification task, units are determined the number of images and the number of human labelers. For example, an image with 3 human labelers counts for 1 * 3 = 3 units. The price for single-label and multi-label classification are the same.
  • For an image bounding box task, units are determined by the number of bounding boxes identified in the images and the number of human labelers. For example, if an image with 2 bounding boxes and 3 human labelers counts for 2 * 3 = 6 units. Images without bounding boxes will not be charged.
  • For an image segmentation/rotated box/polyline/polygon task, units are determined in the same way as a image bounding box task.
  • For a video classification task, units are determined by the video length (every 5 seconds is a price unit) and the number of human labelers. For example, a 25 seconds video with 3 human labelers counts for 25 / 5 * 3 = 15 units. The price for single-label and multi-label classification are the same.
  • For a video object tracking task, unit are determined by the number of objects identified in the video and the number of human labelers. For example, for a video with 2 objects and 3 human labelers, it counts for 2 * 3 = 6 units. Video without objects will not be charged.
  • For a video action recognition task, units are determined in the same way as a video object tracking task.
  • For a text classification task, units are determined by text length (every 50 words is a price unit) and the number of human labelers. For example, one piece of text with 100 words and 3 human labelers counts for 100 / 50 * 3 = 6 units. The price for single-label and multi-label classification is the same.
  • For a text sentiment task, units are determined in the same way as a text classification task.
  • For a text entity extraction task, units are determined by text length (every 50 words is a price unit), the number of entities identified, and the number of human labelers. For example, a piece of text with 100 words, 2 entities identified, and 3 human labelers counts for 100 / 50 * 2 * 3 = 12 units. Text without entities will not be charged.

• For image/video/text classification and text sentiment tasks, human labelers may lose track of classes if the label set size is too large. As a result, we send at most 20 classes to the human labelers at a time. For example, if the label set size of a labeling task is 40, each data item will be sent for human review 40 / 20 = 2 times, and we will charge 2 times of the price (calculated above) accordingly.

• For a labeling task that enables the custom labeler feature, each data item is counted as 1 custom labeler unit.

• For an active learning labeling task for data items with annotations that are generated by models (without a human labeler's help), each data item is counted as 1 active learning unit.

• For an active learning labeling task for data items with annotations that are generated by human labelers, each data item is counted as a regular labeling task as described above.

The table below provides the price per 1,000 units per human labeler, based on the unit listed for each objective. Tier 1 pricing applies to the first 50,000 units per month in each Google Cloud project; Tier 2 pricing applies to the next 950,000 units per month in the project, up to 1,000,000 units. Contact us for pricing above 1,000,000 units per month.

Data type	Objective	Unit	Tier 1	Tier 2
Image	Classification	Image	$35	$25
	Bounding box	Bounding box	$63	$49
	Segmentation	Segment	$870	$850
	Rotated box	Bounding box	$86	$60
	Polygon/polyline	Polygon/Polyline	$257	$180
Video	Classification	5sec video	$86	$60
	Object tracking	Bounding box	$86	$60
	Action recognition	Event in 30sec video	$214	$150
Text	Classification	50 words	$129	$90
	Sentiment	50 words	$200	$140
	Entity extraction	Entity	$86	$60
Active Learning	All	Data item	$80	$56
Custom Labeler	All	Data item	$80	$56

Required use of Cloud Storage

In addition to the costs described in this document, you are required to store data and program files in Cloud Storage buckets during the Vertex AI lifecycle. This storage is subject to the Cloud Storage pricing policy.

Required use of Cloud Storage includes:

• Staging your training application package for custom-trained models.

• Storing your training input data.

• Storing the output of your training jobs. Vertex AI does not require long-term storage of these items. You can remove the files as soon as the operation is complete.

Free operations for managing your resources

The resource management operations provided by AI Platform are available free of charge. The AI Platform quota policy does limit some of these operations.

Resource	Free operations
models	create, get, list, delete
versions	create, get, list, delete, setDefault
jobs	get, list, cancel
operations	get, list, cancel, delete

Google Cloud costs

If you store images to be analyzed in Cloud Storage or use other Google Cloud resources in tandem with Vertex AI, then you will also be billed for the use of those services.

To view your current billing status in the Google Cloud console, including usage and your current bill, see the Billing page. For more details about managing your account, see the Cloud Billing Documentation or Billing and Payments Support.

What's next

• Read the Vertex AI documentation.
• Learn about Vertex AI solutions and use cases.

  Request a custom quote

  With Google Cloud's pay-as-you-go pricing, you only pay for the services you use. Connect with our sales team to get a custom quote for your organization.

  Contact sales