How-to guide

Languages

Set up an agent that speaks more than one language, follows callers when they switch, and see what each provider covers.

What you'll learn

How to tell an agent which languages to speak, how it follows a caller who switches, which voice it uses for each language, and what gets recorded.

Prerequisites

Step 1: List the languages

Open the agent, choose Overview, then Edit, then Model & voice. Under Languages, pick the main language and add the others the agent also speaks. A language that is not in the list can be added by typing its language code, for example sw-KE.

The main language is the one the agent greets callers in, and the one it falls back to when it cannot tell what the caller speaks. How many other languages an agent can list is a limit of your workspace.

Over the API and the SDK the same settings are language, additionalLanguages, languageSwitching, languageOverrides and speechFallback. A chat assistant connected over MCP, and the builder, can set them too.

Step 2: Choose how it follows the caller

ChoiceWhat the agent does
Follow them, in these languagesAnswers in the language the caller speaks, as long as you listed it. This is the default.
Follow them into any languageAnswers in whatever language the caller speaks. The voice may not sound right in a language you have not listed.
Always answer in the main languageUnderstands the other languages and always answers in the main one.

Step 3: Check the voices

Each listed language shows the voice that speaks it. A language uses the agent's voice unless you pick another for it. Press play to hear a voice. Anything the catalog says does not list the language is greyed with the reason, and you can still choose it.

Under the voices, any provider that does not cover a listed language is shown as a plain sentence with a one-tap fix, such as using a transcriber that does hear Gujarati.

What the agent says by itself

Hold, wait, goodbye and apology lines are written for each listed language when you save. Open What the agent says by itself to read them, change any line, or ask for them to be written again. Your own words always win over the written ones.

If the voice company has a problem

Choose whether the call carries on with VoiceLayer's voice, or the agent apologises and ends the call. When Use only your own carrier and keys is on for the workspace, calls never switch to VoiceLayer's voice, and this shows as fixed.

What is recorded

Each call and conversation records the language most of it was held in, and every language it moved through. The call timeline shows when the caller changed language. You can filter Calls and Conversations by language.

What each provider covers

This table is generated from the same catalog the dashboard and the API check against.

ProviderDoesModelLanguages
OpenAISpeaksGPT-4o mini TTSArabic, Chinese, Czech, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Romanian, Russian, Spanish, Swedish, Thai, Turkish, Ukrainian, Vietnamese
OpenAISpeaksTTS-1Arabic, Chinese, Czech, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Romanian, Russian, Spanish, Swedish, Thai, Turkish, Ukrainian, Vietnamese
OpenAISpeaksTTS-1 HDArabic, Chinese, Czech, Danish, Dutch, English, Finnish, French, German, Greek, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Romanian, Russian, Spanish, Swedish, Thai, Turkish, Ukrainian, Vietnamese
OpenAISpeech to speechGPT-4o RealtimeAny language
OpenAISpeech to speechGPT-4o mini RealtimeAny language
DeepgramListensNova-3Bangla, Bulgarian, Catalan, Chinese, Croatian, Czech, Danish, Dutch, English, Estonian, Finnish, French, German, Greek, Gujarati, Hindi, Hungarian, Indonesian, Italian, Japanese, Kannada, Korean, Latvian, Lithuanian, Malay, Marathi, Norwegian, Polish, Portuguese, Romanian, Russian, Slovak, Spanish, Swedish, Tamil, Telugu, Turkish, Ukrainian, Vietnamese
DeepgramListensNova-2Bulgarian, Catalan, Chinese, Czech, Danish, Dutch, English, Finnish, French, German, Greek, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Norwegian, Polish, Portuguese, Romanian, Russian, Spanish, Swedish, Tamil, Thai, Turkish, Ukrainian, Vietnamese
DeepgramSpeaksAura-2English
ElevenLabsSpeaksTurbo v2.5Arabic, Bulgarian, Chinese, Croatian, Czech, Danish, Dutch, English, Filipino, Finnish, French, German, Greek, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Romanian, Russian, Slovak, Spanish, Swedish, Tamil, Turkish, Ukrainian, Vietnamese
ElevenLabsSpeaksMultilingual v2Arabic, Bulgarian, Chinese, Croatian, Czech, Danish, Dutch, English, Filipino, Finnish, French, German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Malay, Polish, Portuguese, Romanian, Russian, Slovak, Spanish, Swedish, Tamil, Turkish, Ukrainian
ElevenLabsSpeaksFlash v2.5Arabic, Bulgarian, Chinese, Croatian, Czech, Danish, Dutch, English, Filipino, Finnish, French, German, Greek, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Malay, Norwegian, Polish, Portuguese, Romanian, Russian, Slovak, Spanish, Swedish, Tamil, Turkish, Ukrainian, Vietnamese
CartesiaListensInk-WhisperArabic, Bangla, Bulgarian, Catalan, Chinese, Croatian, Czech, Danish, Dutch, English, Finnish, French, German, Greek, Gujarati, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Korean, Lithuanian, Malay, Marathi, Norwegian, Polish, Portuguese, Romanian, Russian, Slovak, Spanish, Swedish, Tamil, Thai, Turkish, Ukrainian, Urdu, Vietnamese
CartesiaSpeaksSonic 2Chinese, Dutch, English, French, German, Hindi, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Swedish, Turkish
CartesiaSpeaksSonic TurboChinese, Dutch, English, French, German, Hindi, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Swedish, Turkish
AssemblyAIListensUniversal-StreamingEnglish, French, German, Italian, Portuguese, Spanish
Google (Gemini)Speech to speechGemini 3.1 Flash Live (preview)Arabic, Bangla, Dutch, English, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Marathi, Polish, Portuguese, Romanian, Russian, Spanish, Tamil, Telugu, Thai, Turkish, Ukrainian, Vietnamese
Google (Gemini)Speech to speechGemini 3.8 LiveArabic, Bangla, Dutch, English, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Marathi, Polish, Portuguese, Romanian, Russian, Spanish, Tamil, Telugu, Thai, Turkish, Ukrainian, Vietnamese
SarvamListensSaaras v4Assamese, Bangla, Bodo, Dogri, English, Gujarati, Hindi, Kannada, Kashmiri, Konkani, Maithili, Malayalam, Manipuri, Marathi, Nepali, Odia, Punjabi, Sanskrit, Sindhi, Tamil, Telugu, Urdu
SarvamListensSaaras v3Assamese, Bangla, Bodo, Dogri, English, Gujarati, Hindi, Kannada, Kashmiri, Konkani, Maithili, Malayalam, Manipuri, Marathi, Nepali, Odia, Punjabi, Sanskrit, Sindhi, Tamil, Telugu, Urdu
SarvamSpeaksBulbul v4 FlashBangla, English, Gujarati, Hindi, Kannada, Malayalam, Marathi, Odia, Punjabi, Tamil, Telugu
SarvamSpeaksBulbul v3Bangla, English, Gujarati, Hindi, Kannada, Malayalam, Marathi, Odia, Punjabi, Tamil, Telugu
Azure SpeechListensAzure Speech recognitionArabic, Bangla, Bulgarian, Catalan, Chinese, Croatian, Czech, Danish, Dutch, English, Finnish, French, German, Greek, Gujarati, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Kannada, Korean, Malay, Malayalam, Marathi, Norwegian, Polish, Portuguese, Punjabi, Romanian, Russian, Slovak, Spanish, Swedish, Tamil, Telugu, Thai, Turkish, Ukrainian, Urdu, Vietnamese
Azure SpeechSpeaksAzure neural voicesArabic, Bangla, Bulgarian, Catalan, Chinese, Croatian, Czech, Danish, Dutch, English, Finnish, French, German, Greek, Gujarati, Hebrew, Hindi, Hungarian, Indonesian, Italian, Japanese, Kannada, Korean, Malay, Malayalam, Marathi, Norwegian, Polish, Portuguese, Punjabi, Romanian, Russian, Slovak, Spanish, Swedish, Tamil, Telugu, Thai, Turkish, Ukrainian, Urdu, Vietnamese

On this page