Write Us

Amazon Translate

Amazon Translate is a neural machine translation service from AWS that converts text and documents between 75 languages. Developers call it through an API, CLI or SDK for real-time or batch jobs. It suits apps, support desks and data pipelines on AWS. Standard translation costs $15 per million characters, with two million characters a month free for 12 months.

What is Amazon Translate?

Amazon Translate is a neural machine translation service from Amazon Web Services that converts text and documents between 75 languages. AWS describes it as fast, high quality, affordable and customisable. Developers send text through an API call, the AWS CLI or an SDK, and the service returns the translation in real time or as a batch job for larger collections.

The service runs on deep learning models that AWS keeps updating. Typical uses include translating user-generated content, analysing conversations across languages and supporting multilingual chat, email and help desks. It is a live, actively supported AWS service with pay-as-you-go pricing per character, and it sits alongside other AWS AI services such as Amazon Comprehend, Amazon Transcribe and Amazon Polly.

Key Features

  • Neural machine translation: Amazon Translate uses deep learning models to translate text between 75 supported languages, according to the AWS FAQ. AWS describes the output as fast, high quality and continually improving, and it covers major languages plus many from Africa, Asia and Europe.
  • Real-time translation API: Amazon Translate returns translations from a single API call in real time, with up to 5,000 bytes of text per request. This suits chat messages, product reviews, comments and other short content that users expect to read straight away.
  • Batch translation: Batch jobs translate large document sets stored in Amazon S3, accepting up to 5 GB per call and one million documents per batch. Each document can be up to 20 MB and one million characters, per the official FAQ.
  • Real-time document translation: Amazon Translate also converts whole documents in real time, priced separately for text, HTML and Docx files. Docx costs $30 per million characters, double the standard rate, and this option has no free tier.
  • Office document support: AWS announced support for Office Open XML documents in July 2020. Its machine learning blog shows Word documents, Excel spreadsheets and PowerPoint presentations translated in batch while the file structure stays in place.
  • Custom terminology: Custom terminology lets you fix how brand names, model names and other unique terms are translated. The feature keeps product names consistent across languages without retraining any model.
  • Active Custom Translation: Active Custom Translation adapts output using your own parallel data, meaning example source and target sentence pairs. It costs $60 per million characters, and AWS gives each account 200 GB of parallel data storage free.
  • AWS service integrations: Amazon Translate connects with Amazon S3, DynamoDB, Aurora, Redshift, Comprehend, Transcribe, Polly, Lambda and Glue, according to the developer guide. Teams can chain these services to caption audio, translate stored text or voice the output.

Pros

  • Clear usage-based pricing: Amazon Translate charges a flat $15 per million characters for standard text and batch translation, including spaces. You pay only for what you translate, and there are no seat licences or monthly minimums to manage.
  • Generous first-year allowance: New AWS accounts get two million characters a month free for 12 months on standard text and batch jobs. That covers a lot of testing and small production workloads before any bill arrives.
  • Deep AWS integration: Amazon Translate plugs directly into S3, Comprehend, Transcribe, Polly, Lambda and Glue. Teams already on AWS can add translation to existing pipelines using the same identity, billing and tooling they know.
  • Brand control without training: Custom terminology keeps product and brand names consistent without any machine learning work. Active Custom Translation goes further when you have parallel data, so output can match your house style.
  • Built for volume: Batch jobs process up to one million documents and 5 GB per call. AWS also offers volume discounts above one billion characters a month, which suits enterprises translating large archives or user content.

Cons

  • Developer-focused setup: Amazon Translate needs an AWS account and, for most uses, code or CLI work. Non-technical users looking for a simple translation app will find consumer tools far easier to start with.
  • Free tier expires: The standard free allowance lasts only 12 months, and the Active Custom Translation allowance only two months. After that every character is billed, so long-running hobby projects become paid workloads.
  • Default data use: AWS may store and use submitted text to improve Amazon Translate unless you opt out. Opting out requires an AWS Organizations opt-out policy, an extra step that privacy-conscious teams must remember to configure.
  • Short real-time requests: Real-time calls accept up to 5,000 bytes of text each. Longer content must be split into chunks or sent through batch translation, which adds engineering effort to interactive apps.
  • Premium features cost more: Active Custom Translation costs $60 per million characters, four times the standard rate. Real-time Docx translation costs double and has no free tier, so tailored or document-heavy workloads grow expensive quickly.

Who is Using Amazon Translate?

  • Social and community platforms: Teams translate profiles, comments and posts so users can read content in their own language.
  • Customer support teams: Help desks translate chat, email and ticket messages between agents and customers.
  • Data and analytics teams: Analysts translate multilingual conversations before running sentiment analysis with Amazon Comprehend.
  • Media and eDiscovery teams: Companies translate document repositories, subtitles and search content across languages.

Pricing

Amazon Translate is pay-as-you-go at $15 per million characters for standard translation, with a 12-month free tier for new usage (official AWS prices, October 2026).

Plan Price What you get
Free Tier (standard) Free for 12 months 2 million characters a month of standard text and batch document translation
Standard text and batch $15 per million characters Real-time text, batch documents and real-time text or HTML documents
Real-time Docx $30 per million characters Real-time translation of Word documents, with no free tier
Active Custom Translation $60 per million characters Output tailored with your parallel data, with 500,000 free characters a month for 2 months
Parallel data storage $0.023 per GB a month Applies beyond the free 200 GB per account

Characters are counted including white space, and AWS offers volume discounts for sustained use above one billion characters a month on request. Regional rates and AWS free tier rules can change, so confirm current figures on the AWS pricing page before you budget a project.

What Makes Amazon Translate Unique?

Native links to the rest of AWS set Amazon Translate apart from standalone translation APIs. You can translate files straight from S3, feed results into Comprehend for sentiment analysis, turn speech into multilingual subtitles with Transcribe, and voice the output with Polly, all under one account and one bill. Custom terminology and Active Custom Translation then let you shape the output around your own vocabulary and example translations.

Compatibilities and Integrations

  • Access methods: AWS Management Console, AWS CLI and SDKs in several programming languages.
  • Storage and databases: Amazon S3 for batch jobs, plus text stored in DynamoDB, Aurora and Redshift.
  • AI services: Amazon Comprehend, Amazon Transcribe and Amazon Polly for analysis, captions and speech.
  • Workflow tools: AWS Lambda and AWS Glue for automated translation pipelines.

Amazon Translate Tutorials

A first translation takes only a few minutes once you have an AWS account.

  1. Sign in to the AWS Management Console and open Amazon Translate.
  2. Paste sample text into the real-time translation page and choose the source and target languages.
  3. Upload a custom terminology file if brand or product names must stay fixed.
  4. Call the TranslateText API from the AWS CLI or an SDK inside your application.
  5. Store larger document sets in S3 and start a batch translation job for them.

How We Rated It

Criterion Score Why
Accuracy and Reliability 4.0 Strong output, varies by pair
Ease of Use 3.3 Needs AWS account and code
Functionality and Features 4.2 Real-time, batch and documents
Performance and Speed 4.4 Fast single-call API
Customization and Flexibility 4.4 Terminology and custom translation
Data Privacy and Security 3.7 Opt-out needed for improvement
Support and Resources 4.1 Extensive AWS documentation
Cost-Efficiency 4.2 Low per-character rates
Integration Capabilities 4.7 Native AWS service links
Overall Score 4.1 Average of the nine criteria

Summary

Amazon Translate suits developers and businesses already on AWS that need dependable, high-volume machine translation inside their own apps and data pipelines. Its main limitation is the technical setup, which makes it a poor fit for casual users who just want a translation app. We also recommend checking the data opt-out policy before sending sensitive text. Start with the free tier, test your key language pairs, and add custom terminology early so brand names stay consistent.

FAQ

Frequently Asked Questions

Find answers to the most common questions about our services, reviews, tools, platforms, and more.

Amazon Translate is used to add machine translation to apps, websites and data pipelines. Businesses use it to translate user comments, support chats, emails and document archives, and to prepare multilingual text for sentiment analysis or subtitles within other AWS services.

Amazon Translate costs $15 per million characters for standard and batch translation. Real-time Docx translation costs $30, and Active Custom Translation $60 per million characters. New usage gets two million free characters a month for 12 months, according to AWS.

Amazon Translate supports 75 languages, according to the official AWS FAQ. These include widely spoken languages such as English, Spanish, French, German, Chinese and Japanese, plus many languages from Africa, Asia, Europe and the Americas.

Yes, by default. AWS may store submitted text and use it to maintain and improve Amazon Translate. You can stop content being used for service improvement by applying an AWS Organizations opt-out policy across your accounts before sending sensitive material.

Yes. Batch translation handles document sets in Amazon S3, including Office files, with each document up to 20 MB. Real-time document translation also converts text, HTML and Docx files, with Docx billed at $30 per million characters.

Active Custom Translation is an Amazon Translate feature that tailors output using your parallel data, meaning example sentence pairs. It costs $60 per million characters, includes 500,000 free characters a month for two months, and 200 GB of free data storage.

Amazon Translate accepts up to 5,000 bytes per real-time API call. Batch jobs allow up to 5 GB per call and one million documents, each document capped at 20 MB and one million characters, according to the official AWS FAQ.

Google Cloud Translation API is the closest cloud rival, while Microsoft Azure Translator suits teams on Azure. IBM Watson Natural Language Processing serves IBM customers, and DeepL is worth testing for European language quality. Compare output on your own text first.

Amazon Translate is used to add machine translation to apps, websites and data pipelines. Businesses use it to translate user comments, support chats, emails and document archives, and to prepare multilingual text for sentiment analysis or subtitles within other AWS services.

Amazon Translate costs $15 per million characters for standard and batch translation. Real-time Docx translation costs $30, and Active Custom Translation $60 per million characters. New usage gets two million free characters a month for 12 months, according to AWS.

Amazon Translate supports 75 languages, according to the official AWS FAQ. These include widely spoken languages such as English, Spanish, French, German, Chinese and Japanese, plus many languages from Africa, Asia, Europe and the Americas.

Yes, by default. AWS may store submitted text and use it to maintain and improve Amazon Translate. You can stop content being used for service improvement by applying an AWS Organizations opt-out policy across your accounts before sending sensitive material.

Yes. Batch translation handles document sets in Amazon S3, including Office files, with each document up to 20 MB. Real-time document translation also converts text, HTML and Docx files, with Docx billed at $30 per million characters.

Active Custom Translation is an Amazon Translate feature that tailors output using your parallel data, meaning example sentence pairs. It costs $60 per million characters, includes 500,000 free characters a month for two months, and 200 GB of free data storage.

Amazon Translate accepts up to 5,000 bytes per real-time API call. Batch jobs allow up to 5 GB per call and one million documents, each document capped at 20 MB and one million characters, according to the official AWS FAQ.

Google Cloud Translation API is the closest cloud rival, while Microsoft Azure Translator suits teams on Azure. IBM Watson Natural Language Processing serves IBM customers, and DeepL is worth testing for European language quality. Compare output on your own text first.

Amazon Translate is used to add machine translation to apps, websites and data pipelines. Businesses use it to translate user comments, support chats, emails and document archives, and to prepare multilingual text for sentiment analysis or subtitles within other AWS services.

Amazon Translate costs $15 per million characters for standard and batch translation. Real-time Docx translation costs $30, and Active Custom Translation $60 per million characters. New usage gets two million free characters a month for 12 months, according to AWS.

Amazon Translate supports 75 languages, according to the official AWS FAQ. These include widely spoken languages such as English, Spanish, French, German, Chinese and Japanese, plus many languages from Africa, Asia, Europe and the Americas.

Yes, by default. AWS may store submitted text and use it to maintain and improve Amazon Translate. You can stop content being used for service improvement by applying an AWS Organizations opt-out policy across your accounts before sending sensitive material.

Yes. Batch translation handles document sets in Amazon S3, including Office files, with each document up to 20 MB. Real-time document translation also converts text, HTML and Docx files, with Docx billed at $30 per million characters.

Active Custom Translation is an Amazon Translate feature that tailors output using your parallel data, meaning example sentence pairs. It costs $60 per million characters, includes 500,000 free characters a month for two months, and 200 GB of free data storage.

Amazon Translate accepts up to 5,000 bytes per real-time API call. Batch jobs allow up to 5 GB per call and one million documents, each document capped at 20 MB and one million characters, according to the official AWS FAQ.

Google Cloud Translation API is the closest cloud rival, while Microsoft Azure Translator suits teams on Azure. IBM Watson Natural Language Processing serves IBM customers, and DeepL is worth testing for European language quality. Compare output on your own text first.

Amazon Translate is used to add machine translation to apps, websites and data pipelines. Businesses use it to translate user comments, support chats, emails and document archives, and to prepare multilingual text for sentiment analysis or subtitles within other AWS services.

Amazon Translate costs $15 per million characters for standard and batch translation. Real-time Docx translation costs $30, and Active Custom Translation $60 per million characters. New usage gets two million free characters a month for 12 months, according to AWS.

Amazon Translate supports 75 languages, according to the official AWS FAQ. These include widely spoken languages such as English, Spanish, French, German, Chinese and Japanese, plus many languages from Africa, Asia, Europe and the Americas.

Yes, by default. AWS may store submitted text and use it to maintain and improve Amazon Translate. You can stop content being used for service improvement by applying an AWS Organizations opt-out policy across your accounts before sending sensitive material.

Yes. Batch translation handles document sets in Amazon S3, including Office files, with each document up to 20 MB. Real-time document translation also converts text, HTML and Docx files, with Docx billed at $30 per million characters.

Active Custom Translation is an Amazon Translate feature that tailors output using your parallel data, meaning example sentence pairs. It costs $60 per million characters, includes 500,000 free characters a month for two months, and 200 GB of free data storage.

Amazon Translate accepts up to 5,000 bytes per real-time API call. Batch jobs allow up to 5 GB per call and one million documents, each document capped at 20 MB and one million characters, according to the official AWS FAQ.

Google Cloud Translation API is the closest cloud rival, while Microsoft Azure Translator suits teams on Azure. IBM Watson Natural Language Processing serves IBM customers, and DeepL is worth testing for European language quality. Compare output on your own text first.

Reviews (0)

Write a Review

Write a Review

Your email address will not be published. Required fields are marked *

Your rating *
Select a rating