Today there are 71 different machine translation software to translate your website content and web apps into other languages. We spent 37 hours reviewing the top 70 to find the seven best you can use to localize your content.
What is the Best Machine Translation Software?
- Amazon Translate — best for large-scale businesses
- Crowdin — best for tech startups
- Google Cloud Translation — best for localization experts
- Unbabel — best for inbound marketing focused startups
- Microsoft Azure Translator — best for enterprises with big data
- Systran Translate — best for any business size
- Text United — best for highly technical teams
Our Verdict — Best For Large Scale Businesses
Price: Starts at $15 per million characters
Amazon Translate is enterprise-level machine translation software. Like other Amazon Web Services, it exists at a price point affordable for startups and big companies. It’s perfect for anyone from one-man teams to developers.
Amazon Translate currently supports 75 languages including English, French, German, Spanish, Russian, Hebrew, Portuguese and Chinese.
The Best Part:
Amazon constantly improves its translation quality using industry research data sets. It’s #1 on the list of top-performing machine translation providers in The State of Machine Translation report.
The Worst Part:
You can only use Amazon Translate on an AWS account which may give you issues if you already work with other Computer Automated Translation software (CATs).
Amazon Translate is proven to reduce translation time by 75%. It provides faster communication between global work teams through its synchronous translation process. BMW uses Amazon Translate.
(2 million characters free per month for 12 months)
Best For
Amazon Translate is best for fast-paced, large-scale businesses that need to translate thousands of characters into many different languages.
Top Features
- 200GB of free data storage for every account. Once this storage space is exhausted, you pay $0.023 per GB for data stored past 200GB.
- Pay-as-you-go payment structure allows you to scale your translation work without stress. Once you sign up, you can enjoy the 2 million characters of free translation per month for 12 months, then pay for any extra characters you need.
- Customized translation output allows you to create a glossary for your business, saving editing time. This feature is available on the Active Custom Translation plan.
- High security. The information transmitted between your applications and Amazon Translate enjoys SSL Encryption. The AWS Identity and Access Management (IAM) permissions policy ensures that only verified administrators can access your Amazon Translate account.
- Automatic language identification and cross-channel translation aid prompt communication across multilingual teams.
Pricing
Amazon Translate offers two pricing plans:
- Standard Translation: $15 per million characters, including white spaces (charged monthly). Get 2 million characters translated for free each month for 12 months.
- Active Custom Translation: $60 per million characters, including white spaces (charged monthly). Get 500,000 characters translated for free each month for 2 months.
- Storage (for all plans): $0.023 per GB. You get the first 200GB free.
Try Amazon Translate today with 2 million characters free per month for 12 months.
2. Crowdin
Our Verdict — Best For Tech Startups
Price: Starts at $40/month (billed annually)
Crowdin is a neural machine translation software that positions itself as the translation software for agile project teams working on multilingual products in tech companies.
The Best Part:
- The crowdsourcing workflow incorporates translation into business processes from an early stage: as early as product design. This makes it an easy choice for Web3 project management.
The Worst Part:
- Users on a free plan donate translations to Crowdin's translation memory.
Crowdin's UI/UX is user-friendly and modern. Their user-onboarding documentation is also comprehensive.
Crowdin allows for community-built multilingual products. They went the extra mile to allow you to hire human translators with the right skill set and experience for your project without leaving the platform. Microsoft, Discord, Calendly, and GitHub use Crowdin.
(60,000 hosted words free)
Best For
Crowdin is best for tech startups. Engineering, customer success, marketing, design, and translator teams can collaborate effortlessly with Crowdin.
Top Features
- API integrations with GitHub, GitLab, Bitbucket, Azure Repos, Google Play, Android, Android Studio, iOS, Visual Studio Code, Slack, Figma, MailChimp, Typeform, and a host of other tools.
- Supports over 40 file formats including csv and json.
- Full-stack translation services including proofreaders, human translators, and translation agencies.
- The Source String Editor eliminates the need to wait for developers. It allows you to edit source texts that developers can simply approve.
- Detailed real-time reports so you can see progress on translation and view and manage translation budgets for each stage of your project.
Pricing
Crowdin offers five pricing plans:
- Free Plan: Offers unlimited public projects, 60,000 hosted words, 1 integration, 1 private project and unlimited translators in public projects.
- Pro Plan: Starts at $40 per month. This covers unlimited public projects, 60,000 - 1,060,000 hosted words, unlimited translators, 2 integrations and 2 private projects. This is ideal for managing processes in early-stage startups as translation projects arise.
- Team Plan: Starts at $150 per month. This covers the features from the Pro plan, plus 100,000+ hosted words, room for 5+ managers, 3 integrations and unlimited private projects. On the team plan, you can add managers to set up the project, and API integrations.
- Team+ Plan: Starts at $450 per month. This covers 500,000 hosted words, 5+ managers, unlimited integrations, project groups, invoice billing, advanced workflows and 2FA. You can create teams and set up workflow automation on this package.
- Business Plan: Covers all features from Team+ plan, plus 2,500,000+ hosted words, 10+ managers, permissions granularity, teams, single sign-on (SAML SSO), priority support & SLAs, high volume pricing, custom authorization methods, IP allowlist, and Slack support. You have to contact the sales team to purchase this package.
Crowdin has 2.1 million users and customers in over 160 countries.
Try Crowdin today with 60,000 hosted words for free.
Our Verdict — Best for Localization Experts
Price: Starts at $20 per million characters
Google Translate AI is an AI software that operates on Google's Neural Machine Translation (GNMT) engine to offer paid language translation services on websites and devices. Google Translate is popular as a powerful tool for meeting diverse user-translation needs.
The Best Part:
- Integration with other Google products is easier because all your product and translation work uses the Google Cloud Translation API.
The Worst Part:
- Pricing is complex — it’s based on the type of translation you are doing, how many pages are sent to the API, and how many characters you translate.
Google Cloud Translation is different from free language translation services available on Google, like the Google Translate web app, web browser extension, and Google Translate used with Google Assistant, Google Docs, and Chrome.
At a basic level, Google Cloud Translation makes it possible to do auto-translation on your app and online platform as Google does on their platform.
(500,000 characters per month free)
Best For
Google Cloud Translation is best for localization experts, and translators that work with large enterprises.
Top Features
- AutoML Translation allows translators, and localization experts to build translation models that Google can work with.
- API Key for easy integration with other tools.
- Integration with REST API and gRPC AI provides for faster and lightweight data transmission with increased scalability and bilateral streaming.
- Supports 100+ language pairs including translation between two languages it hasn't even trained on, using Google’s AI.
- Translation Training cuts down translation and proofreading time because the system learns words and phrases you edit and store to translate them automatically in the next occurrence.
Pricing
Google Cloud Translation has two pricing plans:
- Cloud Translation Basic: Starting at $20 per million characters.
- Cloud Translation Advanced: Starting at $20 per million characters. Gives access to TranslateDocument starting at $0.08 per page which allows you to translate PPT, PDF and Docx files.
Try Google Cloud Translation with 500,000 characters per month for free.
4. Unbabel
Our Verdict — Best for Inbound Marketing Focused Startups
Price: Undisclosed
Unbabel combines the speed and scale of machine translation with human editors for fast, high-quality translations that work anywhere.
To get the highest quality translations, Unbabel has a global community of 100,000 translators that are native speakers of 95+ languages.
The Best Part:
- Hands-on support from native speakers on the Unbabel translation team.
The Worst Part:
- Your CRM may be incompatible with Unbabel.
Unbabel only charges for translations sent to end customers. Unbabel supports 95+ language pairs. Adidas, Shopify, FedEx, eBay, and Udemy use Unbabel.
(free inter-team translations)
Best For
Unbabel is best for startups primarily focused on inbound marketing.
Top Features
- Augmented Machine Translation that covers language customization, domain customization, brand customization, and continuous retraining.
- CRM plugin for Zendesk, Oracle, Freshdesk, helpshift, kustomer, Salesforce, and Intercom.
- Unbabel Unit that covers translation of customer success content in email, chat, and FAQ.
- Designated Success and Technical Support rep to ensure you are achieving your objectives with Unbabel.
- Customer service chat translation service for Enterprise customers.
Pricing
Unbabel offers two pricing plans:
- Standard: Up to 30,000 Unbabel units/year (Unbabel’s translation currency), flat free pricing, and standard onboarding.
- Enterprise: Up to 650,000 Unbabel units/year (Unbabel’s translation currency), pricing that decreases with volume, and premium onboarding.
Unbabel prides themselves on not charging per word or per user. Instead, they offer monthly and annual subscription pricing for their standard and enterprise plans based on the content type and volume you want to translate.
To get started, you need to send details about what translation services you need, then book a call with them to chat about pricing.
5. Microsoft Azure Translator
Our Verdict — Best For Enterprises With Big Data
Price: Starts at $10 per million characters
Microsoft Azure Translator is a cloud-based neural machine translation software that renders scalable, multilingual translations to websites, applications, tools and software.
The Best Part:
- Confidential documents are protected under the strict EU requirements.
The Worst Part:
- You may need to write code to integrate Microsoft Azure Translator into your website.
Microsoft Azure Translator is accessible in all parts of the world including China. The service even extends to your company's office tools. KPMG and Volkswagen use Microsoft Azure Translator.
(2 million characters free)
Best For
Microsoft Azure Translator is best for enterprises with big data.
Top Features
- Integrates with Azure Cognitive Services to offer pre-trained models for machine learning and translation API adaptable business processes.
- Uses REST API, providing a "rich functionality set for any developer".
- Bilingual dictionary that provides lots of translation contexts to choose from.
- Customisable deployment mode presents you with choices on how you want to use the Translate service across your existing workflow and tools.
- Compliance certifications such as HIPAA, ISO, SOC, HITRUST, and FedRAMP guarantee your data security.
Pricing
Microsoft Azure Translator has three pricing plans (and two optional add-ons):
- Free Plan: 2 million characters per month with standard text translation, language detection, bilingual dictionary, transliteration, custom translation, and training.
- Standard Plan: Starts at $10 per million characters for text translation and $15 per million characters for document translation.
- Custom Plan: Starts at $40 per million characters for text translation and document translation.
- Training: Starts at $10 per million source + target characters of training data (max. $300/training). This plan covers the number of characters that enter your translation memory as you constantly save linguistic edits peculiar to your brand.
- Custom Model Hosting: $10 per hosted custom translation model per region, per month (if you make a custom translation model). With custom model hosting, you can create a neural machine translation software exclusively engineered to serve your business domain needs on the Cloud.
Volume licensing is available for Enterprise users.
If you commit to a monthly minimum character translation volume, you can get a lower per-character price:
- $2,055 per 250M characters (overage: $8.22 per 1M characters)
- $6,000 per 1000M characters (overage: $6 per 1M characters)
- $22,000 per 4000M characters (overage: $5.50 per 1M characters)
Try Microsoft Azure Translator with your first 2 million characters for free.
6. Systran Translate
Our Verdict — Best for Any Business Size
Price: Starts at €4.99/month
Systran Translate is neural machine translation software with linguistics capacity for 12 domain-specific translations: Legal, Education, Health, Industry, Agriculture, Finance, Business, Technology, IT, Energy, Patent, and Collaboration.
Systran Translate is super fast at processing large volumes of documents without messing up the graphics.
The Best Part:
- Systran is the world's first machine translation software, created in 1968.
The Worst Part:
- There are not many reviews on it despite its popularity in the industry.
Systran has a marketplace where translation experts have trained its models in every language pair and domain. For example, there is a pre-built model for translating English to German in the finance space.
Find your industry domain and language pair to give it a try.
(14-day free trial with 500,000 characters)
Best For
Systran Translate is best for any business size.
Top Features
- Pioneer in machine translation technology and research.
- Translation of document file formats like pdf, Docx, ppt and excel.
- Advanced User Dictionary that allows you to create your word dictionary and contexts.
- Third-party integration with the API is seamless because Systran is built as a REST API.
- Model Studio for translation memories, model training and model standard scoring.
Pricing
Systran Translate has four pricing plans:
- Translate Pro Lite: €4.99 per month/per user (billed annually) for ~150 pages of text/user/month.
- Translate Pro Plus: €13.99 per month/per user (billed annually) for ~600 pages of text/user/month.
- Translate Pro Premium: €31.99 per month/per user (billed annually) for ~2,000 pages of text/user/month.
- Translate Pro Infinite: €64.99 per month (billed annually) for ~4,500 pages of text/user/month. Additional characters cost €8.99 per million. Additional users cost €9.99 per user.
- Translate Pro API: Starting at €8.99 per month for 1 million characters (if you want to integrate Systran Translate into a third-party application)
Try Systran Translate with a 14-day free trial and 500,000 characters
7. TextUnited
Our Verdict — Best For Highly Technical Teams
Price: Starts at €70/month
TextUnited is a standard neural machine translation software management system that serves your translation needs from start to finish.
At the end of each translation process, the process can be automated using the category of content, tone of voice, and context sentiment. Perfect for working around accurate politeness level.
The Best Part:
- White labeling is available in the custom package so you can resell your translations to clients.
The Worst Part:
- It's pricey compared to other machine translation software.
(14-day free trial with 1,000 words)
Best For
TextUnited is best for highly technical teams.
Top Features
- Automated translation production process using the API.
- 30+ file formats can be translated.
- Overlay editor during website translation work.
- Inbuilt Computer Automated Translation (CAT) tool available once you sign in.
- Human translation is available in the custom package.
Pricing
TextUnited has four pricing plans:
- Free: €0/month for up to 1,000 words.
- Basic: €70/month for up to 5,000 words.
- Essential: €250/month for up to 50,000 words.
- Custom: Custom pricing for human translation, white labeling, extended APIs, more users, etc.
You can cancel your subscription anytime.
Try TextUnited with a free 14-day trial and 1,000 words.
Our Top Three Picks
Here’s a quick summary of our top three picks:
- Amazon Translate — best for large-scale businesses
- Crowdin — best for tech startups
- Google Cloud Translation — best for localization experts
Here’s a quick comparison of our top seven picks:
Tool | Entry Offer | Pricing |
|---|---|---|
Amazon Translate | Free plan — 2 million characters free per month for 12 months | Starts at $15/mo |
Crowdin | Free plan — 60,000 hosted words free | Starts at $40/mo |
Google Cloud Translation | Free plan — 500,000 characters per month free | Starts at $20/mo |
Microsoft Azure Translator | Free plan — 2 million characters free | Starts at $10/mo |
Unbabel | None | Undisclosed |
Systran Translate | 14-day free trial | Starts at €4.99/mo/per user |
Text United | 14-day free trial | Starts at €70/mo |
Here are the top 70 machine translation software we considered in this review:
- Acolad
- Alibaba Translate
- Amazon Translate
- AppTek
- Awaken
- Azure Translator
- Babylon Translator
- Baidu Translate
- Bing Microsoft Translator
- Bohemicus
- BytePlus Translate
- ConveyThis
- Crowdin
- Doc Translator
- DeepL
- Eden AI
- Easy Translate
- Fluently
- Geoworkz
- GlobalLink
- Google Translate
- Google Cloud Translation
- Google Cloud Media Translation
- Google Cloud AutoML Translation
- GramTrans
- IBM Watson Translator
- KantanMT
- Language I/O
- Language Studio
- Language Weaver
- Language Wire l
- Lengoo
- Localazy
- Localize
- Lokally
- MemoQ Translator Pro
- Memsource
- Mirai Translate
- MotionPoint
- OneSky
- Pairaphrase
- Papercup
- Pangeanic
- Personal Translator
- Phrase
- POEditor
- Pragma 5.x
- Prisma AI
- PROMT
- Redokun
- Reverso Translation
- RWS Language Cloud
- Slate Desktop
- Smartcat
- Smartling
- Sonix
- Stepes
- Systran Translate
- Taia
- Tauyou
- TextUnited
- Tilde Machine Translation
- Transifex
- Translate
- Translate FX
- Translate Me
- TripLINGO
- Vidby
- Wordscope
- Yando
What is machine translation?
Machine translation is a process where a computer reads text written in one language and then converts it into another language. It's a form of automatic translation.
Where is machine translation used?
Machine translation can be used on any digital platform and in any digital format. From websites to mobile devices, applications, videos, audio, images, and texts.
Why do we need machine translation?
Machine translation is used to convert texts from one language to another so that people who speak different languages can understand what is written. Beyond the need for translation which could be met by human translators, machine translation software makes the process:
- Faster. A human translator will take 5 days to translate 10,000 words. On the other hand, machine translation can translate 100,000 words in seconds.
- Cheaper. The cost of translating 400,000 words which usually requires editing is an average of $40,000. With machine translation software, this can be done for $30.
What are the types of machine translation tools?
The types of machine translation tools could also be referred to as the levels of development machine translation technology has passed through. They are:
- Statistical Machine Translation (SMT) which translates words verbatim.
- Rule-Based Machine Translation (RBMT) translates by comparing grammatical rules.
- Hybrid Machine Translation (HMT) is a combination of statistical machine translation and rule-based machine translation.
- Neural Machine Translation (NMT) uses artificial intelligence to learn languages as a human would, and carry out translations. NMT is a constantly improving translation system.
These are the four main types of machine translation tools. Choosing any one type of machine translation tool is not a topic for consideration. Each is an improvement on the former.
What are the three types of translator software?
The three types of translator software are compilers, interpreters, and assemblers. They control how translation software interprets an instruction before processing it. You can think of them as “the software behind translator software.”
They each represent ways a computer may interpret the language you want to translate (source code) into something readable; an object code (also known as machine code).
- Compilers translate source codes word for word; requiring a lot of storage space.
- Interpreters translate source codes sentence by sentence; taking a long time to process.
- Assemblers assemble the entire source code like mnemonics for complete translation.
Fortunately, you don't need to know all this information to use machine translation software.
What is the difference between a compiler and a translator?
A compiler is one way a computer receives the source code of a language before it processes it for translation. A translator on the other hand is a computer that translates source codes from one language to another.
How accurate is machine translation?
The accuracy of machine translation is not something that can be put to a clear test. Accuracy varies depending on the language pair being translated. For example, the language pairs of countries in the EU region enjoy accurate translations due to their large volume of translated copies of literature within the region. Overall, machine translation is 60%-80% accurate.
Pro tips for using machine translation software
- Hire a professional translator.
Machine translation software does not currently provide accurate and contextualized translations. A professional translator will help you edit and make corrections to the translation work. Machine translation software helps you save time. Professional translators ensure your work fits and passes the right message in the target language.
- Hire a translator with experience in your domain.
Don't hire anyone who can speak and write your target language fluently. Make sure they are experts in the industry or domain you need the documents translated.
- Use Computer Automated Translation (CAT) software for all your editing.
A CAT software stores the trails of the edited work in its Translation Memory (TM). Using a CAT eliminates the need for recurring edits. These edits get saved up in the TM for auto-recall the next time you need to make a similar correction.
- Compile and use a glossary.
Have a glossary of brand-specific terminologies you don't want tampered with in the machine translation process. Uploading this glossary to your machine translation software will reduce translation errors and editing periods over time. 49% of translation errors are due to incorrect terminology. Keep, update and use your brand glossary.
- Insist on a proper onboarding process.
Do not kickstart the translation phase of your business without first ensuring your team is proficient in using the machine translation system you have chosen. A thorough onboarding process will give your team a solid understanding of how the key features and data security system function. An onboarding process will ensure workflow agility, security, and reduced costs.
A great machine translation system offers training as part of their onboarding process. Don't hesitate to reach out to support when you face challenges.
The Bottom Line
Here are our top picks for the best machine translation software to try this year:
- Amazon Translate — best for large-scale businesses
- Crowdin — best for tech startups
- Google Cloud Translation — best for localization experts
- Unbabel — best for inbound marketing-focused startups
- Microsoft Azure Translator — best for enterprises with big data
- Systran Translate — best for any business size
- TextUnited — best for highly technical teams
