Languages and locales supported by Amazon Lex V2
Amazon Lex V2 supports a variety of languages and locales. The languages that are supported, the features that support these languages, and language-specific guidance to improve your bot's performance are provided in this topic.
Supported languages and locales
Amazon Lex V2 supports the following languages and locales.
| Code | Language and locale |
|---|---|
| ar_AE | Gulf Arabic (United Arab Emirates) |
| ca_ES | Catalan (Spain) |
| de_AT | German (Austria) |
| de_DE | German (Germany) |
| en_AU | English (Australia) |
| en_GB | English (UK) |
| en_IN | English (India) |
| en_US | English (US) |
| en_ZA | English (South Africa) |
| es_419 | Spanish (Latin America) |
| es_ES | Spanish (Spain) |
| es_US | Spanish (US) |
| fi_FI | Finnish (Finland) |
| fr_CA | French (Canada) |
| fr_FR | French (France) |
| hi_IN | Hindi (India) |
| it_IT | Italian (Italy) |
| ja_JP | Japanese (Japan) |
| ko_KR | Korean (Korea) |
| nl_NL | Dutch (The Netherlands) |
| no_NO | Norwegian (Norway) |
| pl_PL | Polish (Poland) |
| pt_BR | Portuguese (Brazil) |
| pt_PT | Portuguese (Portugal) |
| sv_SE | Swedish (Sweden) |
| zh_CN | Mandarin (PRC) |
| zh_HK | Cantonese (Hong Kong) |
Languages and locales supported by Amazon Lex V2 features
The following table lists Amazon Lex V2 features that are limited to certain languages and locales. All other Amazon Lex V2 features are supported in all languages and locales.
| Feature | Supported languages and locales |
|---|---|
| AMAZON.AlphaNumeric | All languages and locales except Korean (ko_KR) |
| AMAZON.KendraSearchIntent | English (US) (en_US) |
| Improving speech recognition with a custom vocabulary | English (UK) (en_GB) English (US) (en_US) |
| Automated Chatbot Designer | English (US) (en_US) |
| Region availability | The following languages and locales are not available in the Asia Pacific (Singapore) (ap-southeast-1) and Africa (Cape Town) (ap-south-1) Regions:
|
| Setting intent context | English (US) (en_US) |
| Grammar slot type | English (Australia) (en_AU) English (UK) (en_GB) English (US) (en_US) |
| Using multiple values in a slot | English (US) (en_US) |
| Improving recognition of slot values with runtime hints | English (UK) (en_GB) English (US) (en_US) |
| Capturing slot values with spelling styles | English (Australia) (en_AU) English (UK) (en_GB) English (US) (en_US) |
| Using confidence scores | English (UK) (en_GB) English (US) (en_US) |
Language guidance for Amazon Lex V2
To improve your bot's performance, you should adhere to these guidelines for the following languages.
Arabic
The variety of Arabic that Amazon Lex V2 is trained on is Gulf Arabic. Keep this in mind when providing sample utterances for your bot. Note that Arabic script is written from right to left.
Hindi
Amazon Lex V2 is able to serve Hindi end-users who switch freely between Hindi and English. If you plan to build a bot that supports this language switching, we recommend the following best practices:
-
In the bot definition, write English words in Latin script.
-
At least 50% of your sample utterances should represent language switching within the same sentence. In these utterances, use Devanagari script for Hindi words and Latin script for English words (for example, "मैं ticket book करना चाहता हूं।").
-
If you expect users to communicate with the bot using Hindi words in Latin script or English words in Devanagari script, then you should include examples of Hindi words in Latin script (for example, "mujhe ek ticket book karni hai") and English words in Devanagari script (for example, "मुझे टिकट की बुकिंग में मदद चाहिए") in your sample utterances.
-
If you expect users to communicate with the bot using sentences that are completely in Hindi or completely in English, then you should include sample utterances that are fully in one language (for example, "I want to book a ticket").
Regions
For a list of AWS Regions where Amazon Lex V2 is available, see AWS regions and endpoints in the AWS General Reference.
