Pricing  |  Vertex AI  |  Google Cloud

The costs for Vertex AI remain the same as
they are for the existing products that Vertex AI supersedes. For
example, the cost of training an AutoML image classification model is the same
whether you train it with Vertex AI or with AutoML Vision.

Your bill will report Vertex AI usage under the same SKUs as the existing
products; for example, the cost of training an AutoML image classification
model appears as AutoML Vision on your bill.

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.

AutoML models

For 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 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.

Select a model type below for pricing information.

Image data

Operation Price per node hour (classification) Price per node hour (object detection)
Training $3.15 $3.15
Training (on-device) $4.95 $18.00
Deployment and online prediction $1.25 $1.82
Batch prediction $2.02 $2.02

Video data

Operation Price per node hour
Training $2.94
Predictions $0.42

Tabular data

Operation Price per node hour for classification/regression Price for forecasting
Training $19.32 $19.32 per node hour
Deployment Same price as
predictions for custom-trained models
n/a
Online prediction and online explanation $0.21 n/a
Batch prediction $1.16 No charge during Preview

Text data

Operation Price
Training $3.00 per hour
Deployment $0.05 per hour
Prediction $5.00 per 1,000 text records

Prices for AutoML text prediction requests are computed based on the
number of text records you send for analysis. A text record is
plain text of up to 1,000 Unicode characters (including whitespace and
any markup such as HTML or XML tags).

If the text provided in a prediction request contains more than 1,000
characters, it counts as one text record for each 1,000 characters.
For example, if you send three requests that contain 800, 1,500, and
600 characters respectively, you would be charged for four text records:
one for the first request (800), two for the second request (1,500),
and one for the third request (600).

Custom-trained models

Training

The tables below provide the price per hour of various training
configurations, as well as the number of training
units
used by each configuration. Training
units measure the resource usage of your job; the price per hour of a machine
configuration is the number of training units it uses multiplied by the
region’s cost of training.

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 of each type
of accelerator you use.

The following tables summarize the pricing in each region where
custom training is available.

Americas

The cost of a training job in all
available Americas regions is
$0.49 per hour, per training unit.

Machine types –
price per hour (training units)
n1-standard-4 $0.1900
(0.3878)
n1-standard-8 $0.3800
(0.7755)
n1-standard-16 $0.7600
(1.5510)
n1-standard-32 $1.5200
(3.1020)
n1-standard-64 $3.0400
(6.2041)
n1-standard-96 $4.5600
(9.3061)
n1-highmem-2 $0.1184
(0.2416)
n1-highmem-4 $0.2368
(0.4833)
n1-highmem-8 $0.4736
(0.9665)
n1-highmem-16 $0.9472
(1.9331)
n1-highmem-32 $1.8944
(3.8661)
n1-highmem-64 $3.7888
(7.7322)
n1-highmem-96 $5.6832
(11.5984)
n1-highcpu-16 $0.5672
(1.1576)
n1-highcpu-32 $1.1344
(2.3151)
n1-highcpu-64 $2.2688
(4.6302)
n1-highcpu-96 $3.4020
(6.9429)
a2-highgpu-1g $0.7395
(1.5091)
a2-highgpu-2g $1.4790
(3.0183)
a2-highgpu-4g $2.9579
(6.0365)
a2-highgpu-8g $5.9158
(12.0731)
a2-megagpu-16g $8.7970
(17.9530)
e2-standard-4 $0.1300
(0.2735)
e2-standard-8 $0.2700
(0.5470)
e2-standard-16 $0.5400
(1.0940)
e2-standard-32 $1.0700
(2.1880)
e2-highmem-2 $0.0900
(0.1845)
e2-highmem-4 $0.1800
(0.3689)
e2-highmem-8 $0.3600
(0.7379)
e2-highmem-16 $0.7200
(1.4758)
e2-highcpu-16 $0.4000
(0.8076)
e2-highcpu-32 $0.7900
(1.6153)
n2-standard-4 $0.1900
(0.3964)
n2-standard-8 $0.3900
(0.7928)
n2-standard-16 $0.7800
(1.5856)
n2-standard-32 $1.5500
(3.1712)
n2-standard-48 $2.3300
(4.7568)
n2-standard-64 $3.1100
(6.3424)
n2-standard-80 $3.8800
(7.9280)
n2-highmem-2 $0.1300
(0.2674)
n2-highmem-4 $0.2600
(0.5348)
n2-highmem-8 $0.5200
(1.0695)
n2-highmem-16 $1.0500
(2.1390)
n2-highmem-32 $2.1000
(4.2780)
n2-highmem-48 $3.1400
(6.4170)
n2-highmem-64 $4.1900
(8.5560)
n2-highmem-80 $5.2400
(10.6950)
n2-highcpu-16 $0.5700
(1.1705)
n2-highcpu-32 $1.1500
(2.3411)
n2-highcpu-48 $1.7200
(3.5116)
n2-highcpu-64 $2.2900
(4.6822)
n2-highcpu-80 $2.8700
(5.8527)
c2-standard-4 $0.2100
(0.4260)
c2-standard-8 $0.4200
(0.8519)
c2-standard-16 $0.8300
(1.7038)
c2-standard-30 $1.5700
(3.1947)
c2-standard-60 $3.1300
(6.3894)
Accelerators – price per hour (training units)
NVIDIA_TESLA_A100 $3.1000
(6.3265)
NVIDIA_TESLA_K80 $0.4500
(0.9184)
NVIDIA_TESLA_P4 $0.6000
(1.2245)
NVIDIA_TESLA_P100 $1.4600
(2.9796)
NVIDIA_TESLA_T4 $0.3500
(0.7143)
NVIDIA_TESLA_V100 $2.4800
(5.0612)
Disks – price per GB per hour (training units)
pd-standard $0.0000667
(0.0001)
pd-ssd $0.0002833
(0.0006)

Europe

The cost of a training job in all
available Europe regions is
$0.54 per hour, per training unit.

Machine types –
price per hour (training units)
n1-standard-4 $0.2200
(0.4074)
n1-standard-8 $0.4400
(0.8148)
n1-standard-16 $0.8800
(1.6296)
n1-standard-32 $1.7600
(3.2593)
n1-standard-64 $3.5200
(6.5185)
n1-standard-96 $5.2800
(9.7778)
n1-highmem-2 $0.1370
(0.2537)
n1-highmem-4 $0.2740
(0.5074)
n1-highmem-8 $0.5480
(1.0148)
n1-highmem-16 $1.0960
(2.0296)
n1-highmem-32 $2.1920
(4.0593)
n1-highmem-64 $4.3840
(8.1185)
n1-highmem-96 $6.5760
(12.1778)
n1-highcpu-16 $0.6568
(1.2163)
n1-highcpu-32 $1.3136
(2.4326)
n1-highcpu-64 $2.6272
(4.8652)
n1-highcpu-96 $3.9408
(7.2978)
a2-highgpu-1g $0.8141
(1.5075)
a2-highgpu-2g $1.6282
(3.0151)
a2-highgpu-4g $3.2563
(6.0301)
a2-highgpu-8g $6.5125
(12.0602)
a2-megagpu-16g $9.6840
(17.9334)
e2-standard-4 $0.1700
(0.3061)
e2-standard-8 $0.3300
(0.6123)
e2-standard-16 $0.6600
(1.2246)
e2-standard-32 $1.3200
(2.4491)
e2-highmem-2 $0.1100
(0.2065)
e2-highmem-4 $0.2200
(0.4130)
e2-highmem-8 $0.4500
(0.8259)
e2-highmem-16 $0.8900
(1.6518)
e2-highcpu-16 $0.4900
(0.9041)
e2-highcpu-32 $0.9800
(1.8082)
n2-standard-4 $0.2400
(0.4437)
n2-standard-8 $0.4800
(0.8875)
n2-standard-16 $0.9600
(1.7749)
n2-standard-32 $1.9200
(3.5498)
n2-standard-48 $2.8800
(5.3247)
n2-standard-64 $3.8300
(7.0996)
n2-standard-80 $4.7900
(8.8754)
n2-highmem-2 $0.1600
(0.2993)
n2-highmem-4 $0.3200
(0.5986)
n2-highmem-8 $0.6500
(1.1971)
n2-highmem-16 $1.2900
(2.3943)
n2-highmem-32 $2.5900
(4.7886)
n2-highmem-48 $3.8800
(7.1828)
n2-highmem-64 $5.1700
(9.5771)
n2-highmem-80 $6.4600
(11.9714)
n2-highcpu-16 $0.7100
(1.3104)
n2-highcpu-32 $1.4200
(2.6207)
n2-highcpu-48 $2.1200
(3.9311)
n2-highcpu-64 $2.8300
(5.2415)
n2-highcpu-80 $3.5400
(6.5519)
c2-standard-4 $0.2600
(0.4770)
c2-standard-8 $0.5200
(0.9540)
c2-standard-16 $1.0300
(1.9080)
c2-standard-30 $1.9300
(3.5776)
c2-standard-60 $3.8600
(7.1551)
Accelerators – price per hour (training units)
NVIDIA_TESLA_A100 $3.1000
(5.7407)
NVIDIA_TESLA_K80 $0.4900
(0.9074)
NVIDIA_TESLA_P4 $0.6500
(1.2037)
NVIDIA_TESLA_P100 $1.6000
(2.9630)
NVIDIA_TESLA_T4 $0.3800
(0.7037)
NVIDIA_TESLA_V100 $2.5500
(4.7222)
Disks – price per GB per hour (training units)
pd-standard $0.0000722
(0.0001)
pd-ssd $0.0003194
(0.0006)

Asia Pacific

The cost of a training job in all
available Asia Pacific regions is
$0.54 per hour, per training unit.

Machine types –
price per hour (training units)
n1-standard-4 $0.2200
(0.4074)
n1-standard-8 $0.4400
(0.8148)
n1-standard-16 $0.8800
(1.6296)
n1-standard-32 $1.7600
(3.2593)
n1-standard-64 $3.5200
(6.5185)
n1-standard-96 $5.2800
(9.7778)
n1-highmem-2 $0.1370
(0.2537)
n1-highmem-4 $0.2740
(0.5074)
n1-highmem-8 $0.5480
(1.0148)
n1-highmem-16 $1.0960
(2.0296)
n1-highmem-32 $2.1920
(4.0593)
n1-highmem-64 $4.3840
(8.1185)
n1-highmem-96 $6.5760
(12.1778)
n1-highcpu-16 $0.6568
(1.2163)
n1-highcpu-32 $1.3136
(2.4326)
n1-highcpu-64 $2.6272
(4.8652)
n1-highcpu-96 $3.9408
(7.2978)
a2-highgpu-1g $0.8141
(1.5075)
a2-highgpu-2g $1.6282
(3.0151)
a2-highgpu-4g $3.2563
(6.0301)
a2-highgpu-8g $6.5125
(12.0602)
a2-megagpu-16g $9.6840
(17.9334)
e2-standard-4 $0.1700
(0.3061)
e2-standard-8 $0.3300
(0.6123)
e2-standard-16 $0.6600
(1.2246)
e2-standard-32 $1.3200
(2.4491)
e2-highmem-2 $0.1100
(0.2065)
e2-highmem-4 $0.2200
(0.4130)
e2-highmem-8 $0.4500
(0.8259)
e2-highmem-16 $0.8900
(1.6518)
e2-highcpu-16 $0.4900
(0.9041)
e2-highcpu-32 $0.9800
(1.8082)
n2-standard-4 $0.2400
(0.4437)
n2-standard-8 $0.4800
(0.8875)
n2-standard-16 $0.9600
(1.7749)
n2-standard-32 $1.9200
(3.5498)
n2-standard-48 $2.8800
(5.3247)
n2-standard-64 $3.8300
(7.0996)
n2-standard-80 $4.7900
(8.8754)
n2-highmem-2 $0.1600
(0.2993)
n2-highmem-4 $0.3200
(0.5986)
n2-highmem-8 $0.6500
(1.1971)
n2-highmem-16 $1.2900
(2.3943)
n2-highmem-32 $2.5900
(4.7886)
n2-highmem-48 $3.8800
(7.1828)
n2-highmem-64 $5.1700
(9.5771)
n2-highmem-80 $6.4600
(11.9714)
n2-highcpu-16 $0.7100
(1.3104)
n2-highcpu-32 $1.4200
(2.6207)
n2-highcpu-48 $2.1200
(3.9311)
n2-highcpu-64 $2.8300
(5.2415)
n2-highcpu-80 $3.5400
(6.5519)
c2-standard-4 $0.2600
(0.4770)
c2-standard-8 $0.5200
(0.9540)
c2-standard-16 $1.0300
(1.9080)
c2-standard-30 $1.9300
(3.5776)
c2-standard-60 $3.8600
(7.1551)
Accelerators – price per hour (training units)
NVIDIA_TESLA_A100 $3.1000
(5.7407)
NVIDIA_TESLA_K80 $0.4900
(0.9074)
NVIDIA_TESLA_P4 $0.6500
(1.2037)
NVIDIA_TESLA_P100 $1.6000
(2.9630)
NVIDIA_TESLA_T4 $0.3900
(0.7222)
NVIDIA_TESLA_V100 $2.5500
(4.7222)
Disks – price per GB per hour (training units)
pd-standard $0.0000722
(0.0001)
pd-ssd $0.0003194
(0.0006)

Notes:

  1. All use is subject to the Vertex AI quota policy.
  2. 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:

  • In one-minute increments.
  • At a price per hour as shown in the above table, where the price per hour
    is calculated from a base price and a number of training units, determined
    by the processing configuration you choose when you start your training
    job.
  • With a minimum of 10 minutes per training job.
  • From the moment when resources are provisioned for a job until the job
    finishes.

Prediction and explanation

This table provides 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 a ready state to handle prediction or
explanation requests.

Americas

Prediction
Predictions and explanations
Machine types – price per
node hour
n1-standard-2 Approximations:

us-east4 $0.107
northamerica-northeast1 $0.1046
Other Americas regions $0.095
n1-standard-4 Approximations:

us-east4 $0.214
northamerica-northeast1 $0.2092
Other Americas regions $0.1901
n1-standard-8 Approximations:

us-east4 $0.428
northamerica-northeast1 $0.4183
Other Americas regions $0.3802
n1-standard-16 Approximations:

us-east4 $0.8559
northamerica-northeast1 $0.8367
Other Americas regions $0.7603
n1-standard-32 Approximations:

us-east4 $1.7119
northamerica-northeast1 $1.6733
Other Americas regions $1.5207
n1-highmem-2 Approximations:

us-east4 $0.1332
northamerica-northeast1 $0.1302
Other Americas regions $0.1184
n1-highmem-4 Approximations:

us-east4 $0.2665
northamerica-northeast1 $0.2605
Other Americas regions $0.2367
n1-highmem-8 Approximations:

us-east4 $0.5329
northamerica-northeast1 $0.5209
Other Americas regions $0.4735
n1-highmem-16 Approximations:

us-east4 $1.0659
northamerica-northeast1 $1.0419
Other Americas regions $0.947
n1-highmem-32 Approximations:

us-east4 $2.1317
northamerica-northeast1 $2.0838
Other Americas regions $1.894
n1-highcpu-2 Approximations:

us-east4 $0.0798
northamerica-northeast1 $0.078
Other Americas regions $0.0709
n1-highcpu-4 Approximations:

us-east4 $0.1596
northamerica-northeast1 $0.156
Other Americas regions $0.1417
n1-highcpu-8 Approximations:

us-east4 $0.3192
northamerica-northeast1 $0.312
Other Americas regions $0.2834
n1-highcpu-16 Approximations:

us-east4 $0.6384
northamerica-northeast1 $0.624
Other Americas regions $0.5669
n1-highcpu-32 Approximations:

us-east4 $1.2768
northamerica-northeast1 $1.248
Other Americas regions $1.1338

Europe

Prediction
Predictions and explanations
Machine types – price per
node hour
n1-standard-2 Approximations:

europe-west2 $0.1224
Other Europe regions $0.11
n1-standard-4 Approximations:

europe-west2 $0.2448
Other Europe regions $0.2201
n1-standard-8 Approximations:

europe-west2 $0.4896
Other Europe regions $0.4401
n1-standard-16 Approximations:

europe-west2 $0.9792
Other Europe regions $0.8802
n1-standard-32 Approximations:

europe-west2 $1.9583
Other Europe regions $1.7605
n1-highmem-2 Approximations:

europe-west2 $0.1524
Other Europe regions $0.137
n1-highmem-4 Approximations:

europe-west2 $0.3048
Other Europe regions $0.274
n1-highmem-8 Approximations:

europe-west2 $0.6097
Other Europe regions $0.548
n1-highmem-16 Approximations:

europe-west2 $1.2193
Other Europe regions $1.0959
n1-highmem-32 Approximations:

europe-west2 $2.4386
Other Europe regions $2.1918
n1-highcpu-2 Approximations:

europe-west2 $0.0913
Other Europe regions $0.0821
n1-highcpu-4 Approximations:

europe-west2 $0.1826
Other Europe regions $0.1642
n1-highcpu-8 Approximations:

europe-west2 $0.3651
Other Europe regions $0.3284
n1-highcpu-16 Approximations:

europe-west2 $0.7303
Other Europe regions $0.6567
n1-highcpu-32 Approximations:

europe-west2 $1.4606
Other Europe regions $1.3134

Asia Pacific

Prediction
Predictions and explanations
Machine types – price per
node hour
n1-standard-2 Approximations:

asia-northeast1 $0.1219
asia-southeast1 $0.1172
australia-southeast1 $0.1348
Other Asia Pacific regions $0.11
n1-standard-4 Approximations:

asia-northeast1 $0.2438
asia-southeast1 $0.2344
australia-southeast1 $0.2696
Other Asia Pacific regions $0.2201
n1-standard-8 Approximations:

asia-northeast1 $0.4875
asia-southeast1 $0.4688
australia-southeast1 $0.5392
Other Asia Pacific regions $0.4401
n1-standard-16 Approximations:

asia-northeast1 $0.975
asia-southeast1 $0.9375
australia-southeast1 $1.0784
Other Asia Pacific regions $0.8802
n1-standard-32 Approximations:

asia-northeast1 $1.9501
asia-southeast1 $1.8751
australia-southeast1 $2.1567
Other Asia Pacific regions $1.7605
n1-highmem-2 Approximations:

asia-northeast1 $0.1517
asia-southeast1 $0.1459
australia-southeast1 $0.1679
Other Asia Pacific regions $0.137
n1-highmem-4 Approximations:

asia-northeast1 $0.3034
asia-southeast1 $0.2919
australia-southeast1 $0.3357
Other Asia Pacific regions $0.274
n1-highmem-8 Approximations:

asia-northeast1 $0.6067
asia-southeast1 $0.5837
australia-southeast1 $0.6714
Other Asia Pacific regions $0.548
n1-highmem-16 Approximations:

asia-northeast1 $1.2135
asia-southeast1 $1.1675
australia-southeast1 $1.3428
Other Asia Pacific regions $1.0959
n1-highmem-32 Approximations:

asia-northeast1 $2.4269
asia-southeast1 $2.335
australia-southeast1 $2.6857
Other Asia Pacific regions $2.1918
n1-highcpu-2 Approximations:

asia-northeast1 $0.091
asia-southeast1 $0.0874
australia-southeast1 $0.1005
Other Asia Pacific regions $0.0821
n1-highcpu-4 Approximations:

asia-northeast1 $0.182
asia-southeast1 $0.1748
australia-southeast1 $0.2011
Other Asia Pacific regions $0.1642
n1-highcpu-8 Approximations:

asia-northeast1 $0.364
asia-southeast1 $0.3496
australia-southeast1 $0.4021
Other Asia Pacific regions $0.3284
n1-highcpu-16 Approximations:

asia-northeast1 $0.7279
asia-southeast1 $0.6992
australia-southeast1 $0.8043
Other Asia Pacific regions $0.6567
n1-highcpu-32 Approximations:

asia-northeast1 $1.4558
asia-southeast1 $1.3985
australia-southeast1 $1.6085
Other Asia Pacific regions $1.3134

Each machine type is charged as two
separate SKUs on your Google Cloud bill:

  • vCPU cost, measured in vCPU hours
  • RAM cost, measured in GB hours

The prices for machine typess in the previous table approximate the
total hourly cost for each prediction node of a model version
using that machine type. For example, since an
n1-highcpu-32 machine type includes 32 vCPUs and
28.8 GB of RAM, the hourly pricing per node is equal to 32 vCPU
hours + 28.8 GB hours.

The prices in the previous table are provided to help you
estimate prediction costs. The following table shows the
vCPU and RAM pricing for prediction machine types, which more precisely
reflect the SKUs that you will be charged for:

Americas

Prediction machine type SKUs
vCPU
N. Virginia (us-east4) $0.035605 per vCPU hour
Montréal (northamerica-northeast1) $0.034802 per vCPU hour
Other Americas regions $0.031613 per vCPU hour
RAM
N. Virginia (us-east4) $0.004771 per GB hour
Montréal (northamerica-northeast1) $0.004664 per GB hour
Other Americas regions $0.004242 per GB hour

Europe

Prediction machine type SKUs
vCPU
London (europe-west2) $0.04073 per vCPU hour
Other Europe regions $0.036632 per vCPU hour
RAM
London (europe-west2) $0.005458 per GB hour
Other Europe regions $0.004902 per GB hour

Asia Pacific

Prediction machine type SKUs
vCPU
Tokyo (asia-northeast1) $0.040618 per vCPU hour
Singapore (asia-southeast1) $0.038999 per vCPU hour
Sydney (australia-southeast1) $0.044856 per vCPU hour
Other Asia Pacific regions $0.036632 per vCPU hour
RAM
Tokyo (asia-northeast1) $0.005419 per GB hour
Singapore (asia-southeast1) $0.005226 per GB hour
Sydney (australia-southeast1) $0.006011 per GB hour
Other Asia Pacific regions $0.004902 per GB hour

You can optionally use
GPU
accelerators for prediction. GPUs incur
an additional charge, separate from those described in the
previous table. The following table describes the pricing for each
type of GPU:

Americas

Accelerators – price per hour
NVIDIA_TESLA_K80
Iowa (us-central1) $0.4500
South Carolina (us-east1) $0.4500
NVIDIA_TESLA_P4
Iowa (us-central1) $0.6000
N. Virginia (us-east4) $0.6000
Montréal (northamerica-northeast1) $0.6500
NVIDIA_TESLA_P100
Oregon (us-west1) $1.4600
Iowa (us-central1) $1.4600
South Carolina (us-east1) $1.4600
NVIDIA_TESLA_T4
Oregon (us-west1) $0.3500
Iowa (us-central1) $0.3500
South Carolina (us-east1) $0.3500
NVIDIA_TESLA_V100
Oregon (us-west1) $2.4800
Iowa (us-central1) $2.4800

Europe

Accelerators – price per hour
NVIDIA_TESLA_K80
Belgium (europe-west1) $0.4900
NVIDIA_TESLA_P4
Netherlands (europe-west4) $0.6500
NVIDIA_TESLA_P100
Belgium (europe-west1) $1.6000
NVIDIA_TESLA_T4
London (europe-west2) $0.4100
Netherlands (europe-west4) $0.3800
NVIDIA_TESLA_V100
Netherlands (europe-west4) $2.5500

Asia Pacific

Accelerators – price per hour
NVIDIA_TESLA_K80
Taiwan (asia-east1) $0.4900
NVIDIA_TESLA_P4
Singapore (asia-southeast1) $0.6500
Sydney (australia-southeast1) $0.6500
NVIDIA_TESLA_P100
Taiwan (asia-east1) $1.6000
NVIDIA_TESLA_T4
Tokyo (asia-northeast1) $0.3700
Singapore (asia-southeast1) $0.3700
Seoul (asia-northeast3) $0.3900
NVIDIA_TESLA_V100 Not available

Note that the pricing is per GPU, so if you use multiple GPUs per
prediction node (or if your version scales to use multiple nodes),
then 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.

Vertex Pipelines

Vertex 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 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 Feature Store

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

Operation Price
Online storage $0.25 per GB-month
Offline Storage $0.023 per GB-month
Online Serving $0.94 per node per hour
Batch Export $0.005 per GB

Vertex ML Metadata is free of charge during Preview.

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.
At General Availability, Vertex ML Metadata will charge
$10 per gibibyte (GiB) per month for metadata storage.

Vertex TensorBoard

Vertex TensorBoard is free of charge during Preview. You also pay for
Google Cloud resources you use with Vertex TensorBoard, such as
TensorBoard logs stored in Cloud Storage.

Vertex Vizier

Vertex Vizier is a black-box optimization service inside Vertex AI.
While it is in preview, Vertex Vizier is offered at no charge in all regions
where Vertex Vizier is available.

Vertex Matching Engine

Pricing for Vertex Matching Engine consists of:

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

The per node hour pricing for the Preview release represents a 40% discount over
the price at General Availability.

Region Price per node hour (including 40% discount)
us-central1 $0.7952
asia-southeast1 $0.936

Building and updating indexes is free during Preview. At General Availability,
the price will be $3 per GB of data processed.

Node hour cost will appear on your bill as two separate SKUs:

  • vCPU cost, measured in vCPU hours. Each index hosting node consists of 16 vCPUs.
  • RAM cost, measured in GB hours. Each index hosting node consists of 60GB of RAM.
SKU Region Price per node hour (including 40% discount)
vCPU us-central1 $0.0332
asia-southeast1 $0.0390
RAM us-central1 $0.0044
asia-southeast1 $0.0052

The following tables summarize the pricing of index serving in each region where
matching engine is GA available.

Region Price per node hour
us-central1 $1.0640
us-east1 $1.0640
us-east4 $1.1984
us-west1 $1.0640
asia-southeast1 $1.3126
europe-west1 $1.1715

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.
At General Availability(GA), Vertex Matching Engine will charge
$3 per gibibyte (GiB) of data processed in all regions.

Vertex 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 Model Monitoring.

When you use Vertex 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 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 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.

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 event 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.

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. See the
Google Cloud Pricing Calculator
to determine other costs based on current rates.

To view your current billing status in
the 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

Related Articles

Leave a Reply

Back to top button