With Amazon Rekognition there are 4 different types of usage, each with their own pricing details. Learn more:
Amazon Rekognition Image pricing
Amazon Rekognition Image makes it easy to add image analysis to your applications using proven, highly scalable, deep learning technology that requires no machine learning expertise to use. With Amazon Rekognition Image, you only pay for what you use. There is no up-front commitment or minimum fee. There are two types of costs with Amazon Rekognition Image: the cost for image analysis and the cost for face metadata storage.
Image analysis: Amazon Rekognition Image charges you each time you analyze an image using our APIs. Running multiple APIs against a single image counts as processing multiple images. Usage is billed based on a tiered pricing model tied to volume of images processed per month. The majority of Amazon Rekognition Image APIs are categorized into two groups, Group 1 and Group 2, which have different pricing.
Group1: AssociateFaces, CompareFaces, DisassociateFaces, IndexFaces, SearchFacesbyImage, SearchFaces, SearchUsersByImage, SearchUsers.
Group 2: DetectFaces, DetectModerationLabels, DetectLabels, DetectText, RecognizeCelebrities, DetectProtectiveEquipment APIs.
Other than the APIs listed in Group 1 and Group 2 above, Amazon Rekognition Image also supports Image Properties, which is priced separately from Group 1 and Group 2.
Face metadata storage: To enable face and user search, you will need to store a repository of face metadata objects (face vectors and user vectors) against which Amazon Rekognition can search for matches. Storage charges are applied monthly, and are pro-rated for partial months.
Free Tier
As part of the AWS Free Tier, you can get started with Amazon Rekognition Image for free. The free tier period lasts 12 months from the date of account creation.
Image analysis: During the free tier period you can analyze 1,000 images per month for free each, in Group 1 and Group 2 APIs. Free tier is not offered for Image Properties.
Face metadata storage: During the free tier period, you can store 1,000 face vector objects and 1,000 user vector objects per month for free.
Pricing table
Image Analysis
Face Metadata Storage
Pricing examples
Let's assume your application analyzes 2.5 million images in a month that require label detection. You use Amazon Rekognition’s DetectLabels APIs, to analyze these 2.5 million images.
Total number of images processed with Group 2 APIs (DetectLabels) is 2.5 million.
Cost of processing 2.5 million images with Group 2 APIs
Cost type |
Pricing |
Usage cost |
First 1 million images |
$0.0010 per image |
1,000,000 images X $0.0010/image = $1,000 |
Next 1.5 million images |
$0.0008 per image |
1,500,000 images X $0.0008/image = $1,200 |
Total: $2,200 |
Let's assume your application analyzes 2.5 million images in a month that require image properties for image quality and dominant color detection. You use Amazon Rekognition’s DetectLabels API using IMAGE_PROPERTIES only as input parameter, to analyze these 2.5 million images.
Total number of images processed with Image Properties is 2.5 million.
Cost of processing 2.5 million images with Image Properties
Cost type |
Pricing |
Usage cost |
First 1 million images |
$0.00075 per image |
1,000,000 images X $0.00075/image = $750 |
Next 1.5 million images |
$0.0006 per image |
1,500,000 images X $0.0006/image = $900 |
Total: $1,650 |
Let's assume your application analyzes 1 million images in a month that require both label detection and image properties at the same time. You use Amazon Rekognition’s DetectLabels API with both GENERAL_LABEL and IMAGE_PROPERTIES to analyze these 1 million images.
Since the images need to be processed by both DetectLabels API and Image Properties, you will be charged for both DetectLabels API and Image Properties.
Total number of images processed:
- 1 million for DetectLabels API (GENERAL_LABEL input parameter)
- 1 million for Image Properties (IMAGE_PROPERTIES input parameter)
Total cost are calculated as follows:
API | Cost type | Pricing | Usage cost |
DetectLabels API (GENERAL_LABEL) | First 1 million images | $0.001 per image | 1,000,000 images X $0.001/image = $1,000 |
Image Properties | First 1 million images | $0.00075 per image | 1,000,000 images X $0.00075/image = $750 |
Total: $1,750 |
Let's assume that your application analyzes 3 million images using DetectLabels API. Thereafter, in the same month, you create a collection of 1 million faces using IndexFaces API, run 500,000 AssociateFaces API calls to assign 2 faces to each user, and perform 2.5 million searches in the collection using SearchUsersbyImage API.
Total number of images processed with Group 1 APIs (IndexFaces, AssociateFaces, SearchUsersbyImage) is 4 million.
Total number of images processed with Group 2 APIs (DetectLabels) is 3 million.
Cost of processing 4 million images with Group 1 APIs
Cost type |
Pricing |
Usage cost |
First 1 million images |
$0.0010 per image |
1,000,000 images X $0.0010/image = $1,000. |
Next 3 million images |
$0.0008 per image |
3,000,000 images X 0.0008/image = $2,400 |
Cost of face meta data storage = 1 million face vectors x $0.00001/face vectors per month + 500,000 user vectors x $0.00001/user vectors per month = $15.
Cost of processing 3 million images with Group 2 APIs
Cost type |
Pricing |
Usage cost |
First 1 million images |
$0.0010 per image |
1,000,000 images X $0.0010/image = $1,000. |
Next 3 million images |
$0.0008 per image |
2,000,000 images X 0.0008/image = $1,600 |
Total: $2,600 |
Total cost = $3,400 + $15 + $2,600 = $6015.
Let’s assume your application creates a collection of 10 million faces and 2 million users. After that, in the same month, you perform 55 million user searches in the collection you created, and delete the collection after 15 days. Thereafter you analyze 40 million images to detect labels, and extract text from images.
In this scenario, you first run 10 million IndexFaces API calls to create a face collection, 2 million AssociateFaces API calls to assign 5 faces to each user, and 55 million SearchUsers API calls to perform searches. After that you run 40 million DetectLabels API calls to detect labels, 40 million DetectText API calls to extract text.
Total number of images processed with APIs in Group 1 (IndexFaces, AssociateFaces, SearchUsers) is 67 million.
Total number of images processed with APIs in Group 2 (DetectLabels, DetectText) is 80 million.
Cost of processing 65 million images with Group 1 APIs
Cost type |
Pricing |
Usage cost |
First 1 million images |
$0.0010 per image |
1,000,000 images X $0.0010/image = $1,000. |
Next 4 million images |
$0.0008 per image |
4,000,000 images X 0.0008/image = $3,200 |
Next 30 million images |
$0.0006 per image |
30,000,000 images X 0.0006/image = $18,000 |
Next 32 million images |
$0.0004 per image |
32,000,000 images X 0.0004/image = $12,800 |
Total: $35,000 |
Cost of face meta data storage = 10 million face vectors x $0.00001/face vectors per month + 2 million user vectors x $0.00001/user vectors per month x 0.5 months = $60.
1 million face vectors x $0.00001/face vectors per month + 500,000 user vectors x $0.00001/user vectors per month = $15.
Cost of processing 80 million images with Group 2 APIs
Cost type |
Pricing |
Usage cost |
First 1 million images |
$0.0010 per image |
1,000,000 images X $0.0010/image = $1,000. |
Next 4 million images |
$0.0008 per image |
4,000,000 images X 0.0008/image = $3,200 |
Next 30 million images |
$0.0006 per image |
30,000,000 images X 0.0006/image = $18,000 |
Next 45 million images |
$0.00025 per image |
45,000,000 images X 0.00025/image = $11,250 |
Total: $33,450 |
Total cost = $35,000 + $60 + $33,450 = $68,510.
Amazon Rekognition Video Pricing
Amazon Rekognition Video supports both stored video analysis and real time streaming video events. With Amazon Rekognition Video you pay only for what you use. There are no resources to provision, no upfront costs, or minimum fees.
Amazon Rekognition streaming video events processes video from new or existing Kinesis Video Streams. Rekognition starts processing the Kinesis Video stream only when you send us the notification to start video analysis and can analyze up to 120 seconds of video per event. You pay only for the amount of video processed by Amazon Rekognition. Note: You will pay separately for the Amazon Kinesis Video Streams service.
Amazon Rekognition Stored Video Analysis charges you for videos analyzed from Amazon S3. When running multiple API calls against the same section of the video, you will get charged separately for each API.
Face Metadata Storage: To enable face search, you will need to store a repository of face metadata against which Amazon Rekognition can search for matches. Storage charges are applied monthly, and are pro-rated for partial months.
Pricing table
Streaming Video Events
Stored Video Analysis
Media Analysis
Storage
Feature | Pricing |
Face Metadata Storage |
$0.00001/face metadata per month** |
**Storage charges are applied monthly and are pro-rated for partial months
Free Tier
As part of the AWS Free Tier, you can get started with Amazon Rekognition Video at no cost. The Free Tier lasts 12 months from the date of account creation and includes 60 free minutes of video analysis per month. Amazon Rekognition Video free tier covers Label Detection, Content Moderation, Face Detection, Face Search, Celebrity Recognition, Text Detection and Person Pathing.
Pricing example
Streaming Video Events
Let’s assume you are a connected home camera provider and have 1,000 users. Each user has one camera in their home that streams video when it detects motion. Let’s assume that on average the user has 7 motion events per camera per day. For each motion event, the service provider streams 10 seconds of video to Amazon Rekognition Streaming Video Events.
Monthly charges:
Let’s assume that the AWS region being used for these 1,000 users is US -East. For each motion event, let’s assume that 10 seconds of video is processed, post motion detection, to analyze if there was a person, pet, or package in that video clip. The cost for a minute of video processed is $0.00817.
Total monthly charges will be calculated as follows:
Minutes of video processed per user (each user has one camera) = 10 seconds *7 motion events per camera per day *30 days in a month / 60 = 35 minutes of video processed per user per month.
Total number of minutes of video processed for 1000 users = 35 *1000 = 35,000 minutes
Feature |
Pricing |
Usage cost |
---|---|---|
Label Detection |
$0.00817/min |
35,000 minutes per month * $0.00817/min = $285.95 (ongoing) |
Let’s assume you provide professional video monitoring services for a residential building with 2,000 cameras. Each of these cameras starts streaming video to Rekognition for 10 seconds per event, post motion detection. Let’s assume that on average there are 20 motion events per camera per day.
Monthly charges:
Let’s assume that the AWS region being used for these 2,000 cameras is US -East. For each motion event, let’s assume that 10 seconds of video is processed, post motion detection, to analyze if there was a person in that video clip. The cost for a minute of video processed is $0.00817.
Total monthly charges will be calculated as follows:
Minutes of video processed per user (each user has one camera) = 10 seconds *20 motion events per camera per day *30 days in a month / 60 = 100 minutes of video processed per camera per month.
Total number of minutes of video processed for 2000 cameras = 100*2000 = 200,000 minutes
Feature |
Pricing |
Usage cost |
---|---|---|
Label Detection |
$0.00817/min |
200,000 minutes per month * $0.00817/min = $1,634 (ongoing) |
Stored Video Analysis
Let’s assume that each month your application analyzes 100,000 minutes of videos stored in Amazon S3 through the Label Detection and Shot Detection, and 50,000 minutes through Content Moderation, in the us-east-1 AWS Region.
Feature |
Pricing |
Usage cost |
---|---|---|
Label Detection |
$0.10/min |
100,000 min/month X $0.10/min = $10,000/month (ongoing) |
Shot Detection |
$0.05/min |
100,000 min/month X $0.05/min = $5,000/month (ongoing) |
Content Moderation |
$0.10/min |
50,000 min/month X $0.10/min = $5,000/month (ongoing) |
Amazon Rekognition Custom Labels pricing
With Amazon Rekognition Custom Labels, you can identify the objects and scenes in images that are specific to your business needs. When using Rekognition Custom Labels, there are two types of costs.
Training Hours
There is a cost for each hour of training required to build a custom model with Amazon Rekognition Custom Labels. Amazon Rekognition Custom Labels may run multiple compute resources in parallel to train your model more quickly. This means that the number of hours billed may be more than the number of actual elapsed hours it takes to train the model. The number of training hours needed to train your model depends on many factors such as number of images and number of labels in the training set, and types of machine learning algorithms used to train your model. We have typically seen that 90% of models need less than 24 training hours to train. Models that take more than 72 training hours will be automatically terminated. You will not be charged if your training is automatically terminated.
For example, let’s assume your training started at 4:00pm and ended at 8:30pm, and Amazon Rekognition Custom Labels used two resources in parallel to train your model more quickly. Your total training hours billed would be 9 hours (4.5 hours elapsed time x 2 resources).
Inference Hours
There is a cost for each hour your trained custom model is available to process images. The number of images you can process in a single hour depends on many factors, such as the size of the images processed and the complexity of the custom model. With Amazon Rekognition Custom Labels, you can run multiple compute resources in parallel to process your images more quickly. This means that the number of hours billed may be more than the actual elapsed hours it takes to run your trained custom model.
For example, let’s assume you started your inference with your custom model at 2:00pm and ended at 5:00pm, and you chose to provision two resources in parallel to process your images. Your total inference hours billed would be 6 hours (3 hours elapsed time x 2 resources).
If you expect to process images in batches (e.g. once a day or week, or at scheduled times during the day), you should provision your custom model at scheduled times, process all your images, and then deprovision your resources. If you do not deprovision your resources, you will continue to be charged even if no images are processed.
Free Tier
As part of the AWS Free Tier, you can get started with Amazon Rekognition Custom Labels for free. The Free Tier lasts 12 months from the date of account creation and includes 2 free training hours per month and 1 free inference hour per month.
Pricing table
Pricing examples
Let’s say you are an online vacation website and would like to make it easier for your customers to search your properties. You have identified several relevant labels (Billiard Table, Dining Room, Ocean Facing, etc.) and want to train a custom model to find these features in your images. Let’s assume it takes 10.2 training hours to train your custom model. Let’s also assume you need 0.5 inference hours per day to process all of the images your site receives.
Cost type |
Pricing |
Usage cost |
10.2 training hours |
$1 per hour |
10.2 hours x $1/hour = $10.20 (one-time) |
0.5 inference hours/day |
$4 per hour
|
0.5 hours/day x $4/hour= $2.00/day (ongoing) |
Let’s say you are a manufacturer and want to identify specific machine parts as they pass through an assembly line. You capture images of various machine parts throughout the entire working day. Let’s assume it took 11 training hours to train your model.
Since you have a steady stream of images throughout the day, your model needs to run continuously from 9am to 5pm, a total of 8 hours each day.
Cost type |
Pricing |
Usage cost |
11 training hours |
$1 per hour |
11 hours x $1/hour = $11.00 (one-time) |
8 inference hours/day |
$4 per hour
|
8 hours/day X $4/hour = $32.00/day (ongoing) |
Let’s say you are a social media marketing agency monitoring hundreds of thousands of influencers in order to accurately and quickly identify the best influencers for your clients. You process 450,000 influencer images per day pulled from various social media channels and run them through your Custom Labels model. Let’s assume that processing 440,000 images per day requires roughly 44 inference hours per day. Let’s also assume it took 9 hours to train your model.
At a scale of 440,000 images per day you will need to run at least two inference resources in parallel to meet your needs.
Cost type |
Pricing |
Usage cost |
9 training hours |
$1 per hour |
9 hours x $1/hour = $9.00 (one-time) |
44 inference hours/day |
$4 per hour |
44 hours/day x $4/hour = $180.00/day (ongoing) |
Amazon Rekognition Face Liveness Pricing
Amazon Rekognition Face Liveness verifies that only real users, not bad actors using spoofs, can access your services. Face Liveness analyzes a short selfie video to detect spoofs presented to the camera, such as printed photos, digital photos, digital videos, or 3D masks, as well as spoofs that bypass the camera, such as pre-recorded or deepfake videos. Face Liveness is a fully managed feature that can be easily added to your React web, native iOS, and native Android applications running on most devices with a front-facing camera. No infrastructure management, hardware-specific implementation, or machine learning expertise is required. The feature automatically scales up or down in response to demand, and you only pay for the face liveness checks you perform.
Pricing table
Pricing example
Let’s assume your application performs 400,000 liveness checks in a month and you use the AWS US East (N. Virginia) region for these checks.
Cost of processing first 500,000 checks in US East (N. Virginia) is $0.015 per check
Total checks processed = 400,000
Total cost per month = $0.015 * 400,000 = $6,000
Let’s assume your application performs 1 million liveness checks in a month and you use the AWS US East (N. Virginia) region for these checks.
Cost of processing first 500,000 checks in US East (N. Virginia) is $0.015 per check and next 500,000 checks is $0.0125 per check.
Total checks processed = 1,000,000
Total cost per month = $0.015 * 500,000 + $0.0125 * 500,000 = $13,750
Let’s assume your application performs 4.5 million liveness checks in a month and you use the AWS US East (N. Virginia) region for these checks.
Cost of processing first 500,000 checks in US East (N. Virginia) is $0.015 per check, next 2.5 million checks is $0.0125 per check, and last 1.5 million checks is $0.010 per check.
Total checks processed = 4,500,000
Total cost per month = $0.015 * 500,000 + $0.0125 * 2,500,000 + $0.01 * 1,500,000 = $53,750
Amazon Rekognition Custom Moderation pricing
With Amazon Rekognition Custom Moderation, you can enhance the accuracy of the moderation deep learning model. You can train a custom adapter with as few as 20 annotated images and achieve enhanced accuracy for image analysis tasks related to your specific use case. No infrastructure management, hardware-specific implementation, or machine learning expertise is required.
To create and use an adapter you must:
- Train your adapter with annotated images
- Use your adapter by providing the unique AdapterID to the DetectModerationLabels API for inference
Training cost
There is a cost for each minute of training required to build an adapter with Amazon Rekognition Custom Moderation. Amazon Rekognition Custom Moderation may run multiple compute resources in parallel to train your adapter more quickly. This means that the number of minutes billed may be more than the number of actual elapsed minutes it takes to train the adapter. The number of training minutes needed to train your adapter depends on many factors such as number of images and number of labels in the training set. We have typically seen that 90% of adapters need less than 60 training minutes to train. You will not be charged if your training is automatically terminated.
For example, let’s assume your training started at 4:00pm and ended at 4:35pm, and Amazon Rekognition Custom Moderation used a single resource to train your adapter. Your total training minutes billed would be 35 minutes (35 minutes elapsed time x 1 resource).
As part of the AWS Free Tier, you can get started with Amazon Rekognition Custom Moderation for free. The Free Tier lasts 12 months from the date of account creation and includes 2 free training hours per month to train a custom moderation adapter or a custom labels model.
Inference cost
You will be charged each time you analyze an image using the custom moderation adapter. There is no up-front commitment or minimum fee. Usage is billed based on a tiered pricing model tied to volume of images processed per month.
Pricing table
Pricing example
Let’s assume that it takes you a total of 30 minutes to train your adapter. You analyze 10M images per month with Amazon Rekognition DetectModerationLabels API using your trained adapter. You use the AWS US East (N. Virginia) region for this operation.
Total number of images processed with Custom Moderation Adapter is 10M and total adapter training time is 30 minutes.
Cost type |
Pricing |
Usage cost |
30 training minutes | $5 per hour | 30 minutes X $5/hour = $2.5 (one-time) |
First 1 million images | $0.0012 per image | 1,000,000 images X $0.0012 per image = $1,200 (ongoing) |
Next 4 million images | $0.00096 per image | 4,000,000 images X $0.00096 per image = $3,840 (ongoing) |
Next 5 million images | $0.00072 per image | 5,000,000 images X $0.00072 per image = $3,600 (ongoing) |
Total: $8,642.50 |
One-time cost to train an adapter = $2.5
Ongoing cost to analyze 10M images with the adapter every month = $8,640
Let’s assume it takes you a total of 90 minutes to train your adapter. You analyze 40M images per month with Amazon Rekognition DetectModerationLabels API using your trained adapter. You use the AWS US East (N. Virginia) region for this operation.
Cost type |
Pricing |
Usage cost |
90 training minutes | $5 per hour | 90 minutes X $5/hour = $7.5 (one-time) |
First 1 million images | $0.0012 per image | 1,000,000 images X $0.0012 per image = $1,200 (ongoing) |
Next 4 million images | $0.00096 per image | 4,000,000 images X $0.00096 per image = $3,840 (ongoing) |
Next 30 million images | $0.00072 per image | 30,000,000 images X $0.00072 per image = $21,600 (ongoing) |
Over 30 million images | $0.0003 per image | 5,000,000 images X $0.0003 per image = $1,500 (ongoing) |
Total: $28,147.50 |
One-time cost to train an adapter = $7.5
Ongoing cost to analyze 40M images with the adapter every month = $28,140
Get started with Amazon Rekognition