Release notes for Language Translator

Service API Versioning

API requests in Language Translator v3 require a version parameter that takes a date in the format version=YYYY-MM-DD. Whenever we change the API in a backwards-incompatible way, we release a new minor version of the API.

Send the version parameter with every API request. The service uses the API version for the date you specify, or the most recent version before that date. Don't default to the current date. Instead, specify a date that matches a version that is compatible with your app, and don't change it until your app is ready for a later version.

The current version is 2018-05-01.

13 January 2023

Cloud Foundry deprecation and migration to resource groups

IBM announced the deprecation of IBM Cloud Foundry on 31 May 2022. As of 30 November 2022, new IBM Cloud Foundry applications cannot be created and only existing users are able to deploy applications. IBM Cloud Foundry reaches end of support on 1 June 2023. At that time, any IBM Cloud Foundry application runtime instances running IBM Cloud Foundry applications will be permanently disabled, deprovisioned, and deleted. For more information about the deprecation, see Deprecation of IBM Cloud Foundry.

To continue to use your IBM Cloud applications beyond 1 June 2023, you must migrate to resource groups before that date. Resource groups are conceptually similar to Cloud Foundry spaces. They include several extra benefits, such as finer-grained access control by using IBM Cloud Identity and Access Management (IAM), the ability to connect service instances to apps and service across different regions, and an easy way to view usage per group. For more information about migration, see Migrating Cloud Foundry service instances and apps to a resource group.

The service performs no request logging

The Language Translator service does not log any request data. You cannot opt in to request logging at the request level or at the account level.

18 October 2022

Most of the following changes were made available earlier. They are now documented.

New translation limits for Advanced and Premium plans

The translation limits for service instances on the Advanced and Premium plans have increased. The Translate document method now enforces a document size limit of 50 MB for instances on the Advanced Plan, and 150 MB for instances on the Premium plan. The size limits for the Lite and Standard plans remain 2 MB and 20 MB respectively. For more information, see Before you begin in the topic Translating documents.

New support for additional subtitle file formats

The service now supports additional subtitle formats:

• Apple® iTunes® Timed Text (.itt files)
• Distribution Format Exchange Profile (.dxfp and .xml files)
• Source Code Control (.scc files)
• Synchronized Accessible Media Interchange (.sami and .smi files)
• SubStation Alpha (.ssa files)
• Time Text Markup Language (.ttml files)

For more information about all supported file formats, see Supported file formats.

Identifying file formats for translation

The service accepts two forms of identification for most file formats. You can identify the format of a file that you send for translation in one of two ways:

• By specifying the appropriate file extension for the format.
• By specifying the content type (MIME type) of the format as the type of the file parameter.

The documentation lists the valid file extensions and content types for each supported format. In most cases, specifying the correct file extension is preferred because it can eliminate ambiguity and is simpler. For subtitles, the documentation makes clear where either the file extension or the content type is needed. For more information about all file formats, their file extensions and content types, and how and when to specify the file extension or content type, see Supported file formats.

Documentation updates for bidirectional translation

The documentation now states that the service correctly translates from and to bidirectional languages that are written left-to-right and right-to-left (for example, Arabic, Hebrew, and Urdu). This is not new functionality, but the documentation had failed to emphasize the information.

5 August 2022

The following changes were made available on 1 August 2021. They are now documented.

Most document formats for translation are now generally available

Most supported file formats for document translation are now generally available (GA). In addition, subtitle formats for documentation translation are also GA. The PDF format remains experimental. For more information, see Supported file formats.

15 June 2022

The following changes were made available on 1 November 2021. They are now documented.

New customizable translation models

Added new customizable translation models for the following languages:

• English to Kannada (en-kn)
• English to Marathi (en-mr)
• English to Punjabi (Indian) (en-pa)
• English to Punjabi (Pakistani) (en-pa-PK)
• Kannada to English (kn-en)
• Marathi to English (mr-en)
• Punjabi (Indian) to English (pa-en)
• Punjabi (Pakistani) to English (pa-PK-en)

The Kannada language is new; it is not identifiable. All of the languages except for Kannada already existed and are identifiable. For more information, see Supported languages for translation.

Improved customizable translation models

Updated the existing customizable translation models for the following languages for improved translation:

• Bengali to English (bn-en)
• Malayalam to English (ml-en)
• Tamil to English (ta-en)
• Telugu to English (te-en)

For more information, see Supported languages for translation.

23 October 2020

New customizable translation models

Added improved versions of the following customizable translation models for Korean:

• English to Korean (en-ko)
• Korean to English (ko-en)

The models are improved to preserve markup in the translation if present in the source.

16 October 2020

New customizable translation models

Added the following customizable translation models for Basque:

• Basque to Spanish (eu-es)
• Spanish to Basque (es-eu)

For more information, see List of supported languages.

25 September 2020

New translation models

Added the following translation models for Montenegrin:

• Montenegrin to English (cnr-en)
• English to Montenegrin (en-cnr)

Montenegrin cannot be identified by the service's language identification feature. (Bosnian, Croatian, and Montenegrin translation support is based on the Latin alphabet. Of the three very similar languages, only Croatian is identifiable.) The cnr-en and en-cnr translation models are not customizable at this time. For more information, see List of supported languages.

16 September 2020

Beta support for subtitle (caption) documents

Added beta support for translation of the following subtitle (caption) document formats:

• SubRip: .srt
• SubViewer: .sbv
• DirectVobSub or VSFilter: .sub
• MicroDVD: .sub
• WebVTT: .vtt

These textual formats contain the transcript of a sound track or video source. For more information about the formats and the characteristics of subtitle translation, see Supported subtitle formats.

14 August 2020

New translation models

Added the following translation models:

• Bosnian to English (bs-en) and English to Bosnian (en-bs). Bosnian cannot be identified by the service's language identification feature.
• Canadian French to English (fr-CA-en) and English to Canadian French (en-fr-CA). Canadian French cannot be identified by the service's language identification feature.
• English to Welsh (en-cy). (The cy-en model was already available.)
• Serbian to English (sr-en) and English to Serbian (en-sr). Serbian translation support is based on the Cyrillic alphabet. (Bosnian and Croatian translation support is based on the Latin alphabet.)

For more information, see List of supported languages.

31 July 2020

New method for listing supported languages

Added the List supported languages (GET /v3/languages) method to retrieve the list of supported languages for translation. The method returns a complete list of all supported languages, sorted by language code (for example, af, ar). In addition to basic information about each language, the response indicates whether the language is supported_as_source for translation and supported_as_target for translation. It also lists whether the language is identifiable.

For more information, see Listing supported languages for translation.

14 July 2020

New supported file types for customization

Added support for new file types for customization: CSV, TSV, XLSX (Microsoft Excel), XLIFF, and JSON. For more information, see Supported document formats for training data.

New translation models

Added Welsh-to-English (cy-en) support in translation, and added support for Welsh (cy) in language detection. For more information, see List of supported languages.

11 June 2020

New customizable translation models

Added the following customizable translation models for Ukrainian:

• Ukrainian to English (uk-en)
• English to Ukrainian (en-uk)

For more information, see List of supported languages.

Improved translation between Spanish and English

Improved results for the translation models between Spanish and English:

• Spanish to English (es-en)
• English to Spanish (en-es)

Improved document translation for Microsoft PowerPoint

Improved table handling with document translation for Microsoft PowerPoint.

28 May 2020

Expanded language support for translation and identification

Expanded language support:

• Support for translating the following languages is now available:
  • Nepali (ne)
  • Sinhala (si)
• Support for identifying the following languages is now available:
  • Burmese (my)
  • Lao (lo)
  • Marathi (mr)
  • Nepali (ne)
  • Punjabi (Shahmukhi script, Pakistan) (pa-PK)
  • Sinhala (si)
  • Tagalog (tl)

Improved translation results for Catalan and Chinese

Improved results when translating to and from the following languages:

• Catalan (ca)
• Chinese (Simplified) (zh)

New translation limits for Lite plan

For service instances on the Lite plan, the Translate document method now has a document size limit of 2 MB. The document size limit for Standard, Advanced, and Premium plans remains 20 MB.

22 April 2020

Expanded language translation

The languages listed in List of supported languages can be translated to any other language in the list, with the exception of Catalan.

Expanded language support for translation

Support for translating the following languages is now available:

• Bengali (bn)
• Gujarati (gu)
• Malayalam (ml)
• Maltese (mt)
• Tamil (ta)
• Telugu (te)

Improved translation for English and Hindi

Improved translation quality for translation between English and Hindi.

Improved translation for Microsoft document types

Improved document translation performance for Microsoft Word, PowerPoint, and Excel documents.

28 February 2020

Automatic source language detection in translation requests

Added automatic source language detection in Translate requests. When you specify a target language without specifying a source language or translation model, the service attempts to identify the source language and continues with the translation. In the results, the service returns the identified language and a score indicating the confidence in the identification.

30 January 2020

New translation models

New translation models are now available:

• English to and from Latvian (en-lv and lv-en)
• English to and from Urdu (en-ur and ur-en)
• English to and from Vietnamese (en-vi and vi-en)

For more information, see List of supported languages.

Improved translation for Microsoft document types

Made general improvements for translating Microsoft Word and PowerPoint documents.

12 December 2019

Full support for IBM Cloud IAM

The Language Translator service now supports the full implementation of IBM Cloud Identity and Access Management (IAM). API keys for Watson services are no longer limited to a single service instance. You can create access policies and API keys that apply to more than one service, and you can grant access between services.

• To support this change, the API service endpoints use a different domain and include the service instance ID. The pattern is api.{location}.{offering}.watson.cloud.ibm.com/instances/{instance_id}.

  Example URL for an instance hosted in the Dallas location: api.us-south.language-translator.watson.cloud.ibm.com/instances/6bbda3b3-d572-45e1-8c54-22d6ed9e52c2

  The previous public endpoint domain was watsonplatform.net.

  For more information about the URLs, see the API & SDK reference.

  These URLs do not introduce a breaking change. The new URLs work both for your existing service instances and for new instances. The original URLs continue to work on your existing service instances for at least one year (until December 2020).

• For more information about IAM, see Authenticating to Watson services.

New network and data security features

Support for the following new network and data security features:

• Support for private network endpoints
  • Users of Premium plans can create private network endpoints to connect to Language Translator over a private network. Connections to private network endpoints do not require public internet access. For more information, see Public and private network endpoints.
• Support for data encryption with customer-managed keys
  • Users of new Premium and Dedicated instances can integrate IBM® Key Protect for IBM Cloud® with Language Translator to encrypt their data and manage encryption keys. For more information, see Protecting sensitive information in your Watson service.

13 November 2019

New Seoul location

You can now create Language Translator instances in the Seoul, South Korea, location. As with other locations, the IBM Cloud Seoul location uses token-based Identity and Access Management (IAM) authentication.

11 November 2019

Customization and identification improvements

Improved custom model training time. Also, Improved language identification accuracy for English text and for text with more than 100 words.

New translation models

New translation models are now available.

• English to and from Indonesian (en-id and id-en)
• English to and from Irish (en-ga and ga-en)
• English to and from Lithuanian (en-lt and lt-en)
• English to and from Malay (en-ms and ms-en)
• English to and from Thai (en-th and th-en)

For more information, see List of supported languages. Note that Indonesian cannot be identified by the service's language identification feature.

21 August 2019

New translation models

New translation models are now available:

• English to and from Bulgarian (en-bg and bg-en)
• English to and from Croatian (en-hr and hr-en)
• English to and from Estonian (en-et and et-en)
• English to and from Romanian (en-ro and ro-en)
• English to and from Slovak (en-sk and sk-en)
• English to and from Slovenian (en-sl and sl-en)

New identifiable languages

The following languages can now be identified by the service:

• Catalan (ca)
• Croatian (hr)
• Irish (ga)
• Malay (ms)
• Maltese (mt)
• Serbian (sr)
• Slovenian (sl)
• Thai (th)

14 June 2019

New translation models

New translation models are now available for English and Greek:

• English to Greek (en-el)
• Greek to English (el-en)

13 June 2019

New translation models

New translation models are now available for English and Hebrew:

• English to Hebrew (en-he)
• Hebrew to English (he-en)

21 March 2019

Changes to service credential information

From March 21 2019, you will see only service credential information associated with the role that has been assigned to your IBM Cloud account. For example, if you have assigned a reader role, any writer or higher levels of service credentials will not be visible.

This change does not affect API access for users or applications with existing service key credentials. Only the viewing of credentials within IBM Cloud is affected.

For more information about service keys and user roles, see Authenticating to Watson services.

14 December 2018

New London location

You can now create Language Translator service instances in the IBM Cloud London location.

16 November 2018

New beta support for document translation

Translating documents is now available through new API endpoints. Submit a Microsoft Office document, PDF, or other document with a supported file format, and Language Translator will provide a translated copy that preserves the original formatting. Supported file formats include .doc, .ppt, .pdf, and more.

New translation models

New translation models for Hungarian are now available:

• Hungarian to English (hu-en)
• English to Hungarian (en-hu)

8 November 2018

New Tokyo location

You can now create Language Translator service instances in the IBM Cloud Tokyo location.

9 August 2018

New translation models

New translation models for Norwegian Bokmål are now available:

• Norwegian Bokmål to English (nb-en)
• English to Norwegian Bokmål (en-nb)

27 June 2018

New translation models

New translation models that feature six new languages are now available:

• Catalan
  • Catalan to Spanish (ca-es)
  • Spanish to Catalan (es-ca)
• Czech
  • Czech to English (cs-en)
  • English to Czech (en-cs)
• Danish
  • Danish to English (da-en)
  • English to Danish (en-da)
• Finnish
  • Finnish to English (fi-en)
  • English to Finnish (en-fi)
• Hindi
  • Hindi to English (hi-en)
  • English to Hindi (en-hi)
• Swedish
  • Swedish to English (sv-en)
  • English to Swedish (en-sv)

15 June 2018

New support for IBM Cloud IAM

As of 15 June 2018, new service instances created in the Germany and US South regions use Identity and Access Management (IAM) authentication.

New service instances that you create in Germany and US South will not be compatible with Language Translator v2. If you use Language Translator v2 and are planning to use new service instances in your application, migrate to the v3 API.

12 June 2018

Version 3 of the service is now available

Language Translator v3 is now available. The v2 Language Translator API will no longer be available after July 31, 2018. To take advantage of the latest service enhancements, migrate to the v3 API.

New features in version 3

Version 3 includes the following new features:

• Language Translator API v3 comes with Neural Machine Translation (NMT) models that offer significantly improved translation results. All NMT models are now available for customization.
• Use custom models as base models for forced glossary customization.
• API v3 is a simplified, all-JSON subset of the retired API v2.
• Introduces API version dates to give developers the freedom to adopt future API changes at their own pace.

Breaking changes in version 3

Version 3 includes the following breaking changes:

• Mandatory version date for all API endpoints: API v3 requests require a version date query parameter of the form version=YYYY-MM-DD. The latest API version is version=2018-05-01.
• Simplified API:
  • The Translate and Identify methods do not offer the option to return plain text responses in v3. The methods return only JSON responses.
  • GET /translate and GET /identify methods are not supported in v3. Use the POST /translate and POST /identify methods instead.
• Monolingual corpus customization is not supported in v3.
• Creating custom models with both a parallel corpus and forced glossary now needs to be done in two API calls. First, customize the model with a parallel corpus. After the custom model has finished training, customize it again with the forced glossary. This change allows you to use a custom model trained with a parallel corpus as a base for forced glossary customization.
• Specialized patent and conversation domain models are not available in the v3 API. The translation quality provided by the NMT models in the patent and conversation domains is generally improved compared to the older specialized models.
• Error object keys have been renamed so that they are consistent with other Watson APIs. error_code has been renamed to code, and error_message has been renamed to error.

New support for IBM Cloud IAM

As of 12 June 2018, new service instances created in the Sydney and US East regions use Identity and Access Management (IAM) authentication.

12 January 2018

New Neural Machine Translation Models

New Neural Machine Translation (NMT) models are available to preview. You can try NMT models for the following language pairs.

• English to and from: Arabic, Chinese, Dutch, French, German, Italian, Japanese, Korean, Polish, Portuguese (Brazilian), Russian, Spanish, and Turkish
• French to and from: German, Spanish
• German to and from: Italian

The NMT models and the syntax for using them are subject to change during the preview period. Currently, NMT models do not support corpus customization. Only forced glossary customization is supported.

To use an NMT preview model to translate, specify the header X-Watson-Technology-Preview:2017-07-01 along with the character codes for the source and target languages of the model you want to use. The following example shows how to translate English to Spanish with an NMT preview model.

curl -X POST --user {username}:{password} \
-H "Accept: application/json" \
-H "X-Watson-Technology-Preview:2017-07-01" \
-d '{"text":"Hello!","source":"en","target":"es"}' \
"https://gateway.watsonplatform.net/language-translator/api/v2/translate"

15 December 2016

New translation models

New translation models were added: English to and from Japanese

15 November 2016

Beta tooling no longer available or supported

The beta tooling that was previously available for the Language Translator service is no longer available or supported. Contact your sales representative or customer support for information about how to use the Language Translator API to accomplish the tasks supported by the Language Translator tool.

1 September 2016

Service rebranding

The IBM Watson™ Language Translation service was rebranded as the Language Translator service.

22 March 2016

New language support

Support for additional languages was added: English to and from Italian, and Spanish to and from French.

3 December 2015

Customization no longer available with Standard plan

As of January 15, 2016, all customization capabilities within the Standard plan are discontinued. Applications that do not use customization features do not need to change, as the Standard plan remains active for all API calls unrelated to customization or customized models. To use the GA customization features (the Trainable plan) of the Language Translator service with an IBM Cloud® application that uses an earlier instance of the service, complete the following steps:

1. Create a new Watson Language Translator instance and specify the GA "Trainable" plan.
2. Bind the new "Trainable" instance of the service to your app in IBM Cloud.
3. Gather the data that was used to initially create the customized models.
4. Upload the training data to create new customized models on the "Trainable" instance.
5. In your app, point the "ModelID" field to the new customized models.
6. Unbind the earlier service from your app in IBM Cloud, and then delete it.

6 November 2015

Beta tooling available

The beta Language Translator tool is released. The tool is a web application that provides a graphical user interface to manage and train models for more accurate machine translation. You can create projects, upload training data, train custom models, and translate text.

1 December 2014

Beta machine translation and language identification APIs upgraded

The beta Machine Translator and beta Language Identification APIs have been upgraded and combined into the Language Translator API. To immediately start using the new service, understand and update your code to reflect these changes:

• New model_id parameter: In the beta API, you defined the sid parameter to select the model to use for translation. In this version, the sid parameter is renamed to model_id parameter. To retrieve the model_id allowed values, use the GET/language translator/api/v2/models operation. This returns a list of all models and their corresponding model_id values.
• Language pair support: Instead of selecting a model_id, you can now specify a source and target language instead, and the model will default to the one that's trained on the general news domain.
• JSON request body support: When making a POST translation request, you can now make the request as a JSON submission. The JSON formatting allows you to submit multiple paragraphs for translation, instead of just a single piece of text in the form submission format.
• JSON response body support: The translation request returns support JSON formatting as well as plain text formatting. The JSON format allows support for the translated words to be returned in multiple paragraphs instead of a single piece of text.
• Accept header support: The accept header can now be used to define the format of the response in all of the operations (text/plain or application/json).
• Language Identification support: Language identification methods have been added to this API. This allows you to identify the language of the input texts, and lists all supported languages that can be detected by the API.

New API authentication process

The Language Translator service has a new API authentication process for service instances that are hosted in the following locations:

• Dallas as of June 15, 2018
• Frankfurt as of June 15, 2018
• London
• Seoul
• Sydney as of June 12, 2018
• Tokyo
• Washington, DC as of June 12, 2018

IBM Cloud is migrating to token-based Identity and Access Management (IAM) authentication. With some service instances, you authenticate to the API by using IAM.

IBM Cloud IAM available for service instances

IBM Cloud Identity and Access Management (IAM) is available for new and existing service instances as follows:

• For new service instances created in the locations indicated previously, you use IAM for authentication. You can pass either a bearer token or an API key. Tokens support authenticated requests without embedding service credentials in every call. API keys use basic authentication.

  When you use any of the Watson SDKs, you can pass the API key and let the SDK manage the lifecycle of the tokens. For more information and examples, see Authentication in the API & SDK reference.

• For existing service instances that you created before the indicated date, you continue to authenticate by providing the username and password for the service instance. Eventually, you will need to migrate these service instances to IAM authentication. Updates will be provided about migration process and dates. For more information about migration, see Migrating Watson services from Cloud Foundry.

To find out which authentication to use, view the service credentials by clicking the service instance on the IBM Cloud resources page.