Text-to-speech API for language mixing and multilingual speech
Generate mixed-language speech in one request. Switch languages mid-sentence, pronounce foreign names and phrases naturally, and keep the same voice across 60+ languages without splitting text or stitching audio.
$0.70 per generated hour.
Trusted by teams building global voice products
Mixed-language TTS without splitting or stitching
Real-world text rarely stays inside one language. Customer names, addresses, brands, technical terms, quotes, and entire phrases often come from somewhere else. Soniox Text-to-Speech API speaks mixed-language input as one continuous utterance, so your application does not need to split text by language and combine separate audio clips.
Switch languages naturally mid-sentence
Send text containing multiple languages in the same request. Soniox follows the language changes while preserving one continuous voice and delivery.
Your reservation is confirmed. Cuando llegues al hotel, muestra este código en recepción: ES-4928.
The meeting has been moved to Tuesday. Merci de confirmer votre disponibilité avant la fin de la journée.
To enable the new feature, open Settings and select 음성 복제, then tap Continue.
Your order is ready for pickup. 店頭で注文番号 A-7392 をお見せください.
Handle code-switching in one utterance
Build speech for bilingual and multilingual content where languages change naturally inside a sentence or response. Keep the full utterance together instead of routing each language segment through a separate TTS request.
Send mixed text
Send the complete multilingual sentence, message, or response to Soniox Text-to-Speech API as one input.
Keep one request
Set the overall delivery language without splitting every foreign name, phrase, or language change into another request.
Generate one voice
Receive continuous speech that preserves the same speaker identity as the utterance moves between languages.
Pronounce foreign names, places, and terms naturally
Keep names, addresses, brands, product terms, and borrowed words inside the surrounding text. Soniox pronounces them naturally without requiring you to change the language setting for every embedded term.
Your meeting with Nguyễn Minh Anh is scheduled for 3 p.m. at our office near Place de la République in Paris.
Open the Soniox Developer Console, select Text-to-Speech, and set the output to 24-kilohertz PCM before deploying on Kubernetes.
Your appointment with Siobhan O’Connor is confirmed.
Nguyễn Minh Anh will be joining the meeting shortly.
Our Asian engineering office is located in Guangzhou.
Your hotel is located near the historic center of Aix-en-Provence.
Set the delivery language, not every language switch
The language field controls the overall delivery and applies a light accent bias to the voice. Embedded words and phrases from other languages can still be pronounced naturally inside the same utterance.
English delivery
English is the primary delivery language, so the utterance follows English prosody. The Spanish street name and title remain part of the same request and are pronounced naturally.
{
"model": "tts-rt-v2",
"language": "en",
"voice": "Adrian",
"text": "Please deliver the package to Calle de Alcalá 45 in Madrid, attention Señora García."
}Spanish delivery
The text stays exactly the same, but Spanish now controls the overall delivery. The same voice speaks the complete mixed-language utterance with Spanish driving its prosody.
{
"model": "tts-rt-v2",
"language": "es",
"voice": "Adrian",
"text": "Please deliver the package to Calle de Alcalá 45 in Madrid, attention Señora García."
}Use the main language of the text
Choose this when most of the utterance is in one language and names, addresses, technical terms, or shorter phrases come from other languages.
Or choose the desired delivery language
Choose this when you want the overall voice delivery to follow a particular language community even when the input contains substantial text from another language.
Keep the same voice across languages
Language mixing should not sound like several different speech systems stitched together. Every Soniox voice works across all 60+ supported languages, preserving one speaker identity as the language changes.
One voice across 60+ languages
Use the same built-in voice for multilingual applications, localized products, and mixed-language speech while maintaining consistent voice identity and quality.
Clone once, speak every supported language
Create a cloned voice from up to 20 seconds of audio and use that same voice across all 60+ supported languages, including mixed-language utterances.
Mixed-language speech still needs every detail right
Multilingual applications often combine language changes with names, numbers, contact details, addresses, and other structured information. Soniox keeps those details clear inside the same spoken response.
Speak numbers, addresses, and identifiers clearly
Phone numbers, email addresses, verification codes, prices, and postal addresses remain clear and precise alongside multilingual text.
You can reach our support team at +1 415 682 9074.
Send the completed form to alex.chen+support@example.com.
Your verification code is 7Q4M9B. I repeat: 7Q4M9B.
Your delivery is going to 1427 North St. Andrews Place, Apartment 6B, Los Angeles, California 90028.
Your total is $1,284.37, including tax and delivery.
Language mixing for real-time voice applications
Use mixed-language TTS in both single requests and streaming sessions. Generate multilingual speech progressively while keeping one continuous voice and precise control over what has already been spoken.
Stream mixed-language speech as text arrives
Send text progressively and begin playback before the complete message is available, including responses that contain multiple languages, foreign names, and embedded phrases.
Keep timing aligned through language changes
Character-level timestamps let your application synchronize generated speech with text, stop cleanly during interruptions, and track exactly what has already been spoken.
Estimate your multilingual text-to-speech cost
Set your expected generated speech volume to estimate the cost of using Soniox Text-to-Speech API for multilingual and mixed-language applications.
Pricing calculator
Stop overpaying for speech AI
1,000 hours of speech / month
Pricing assumptions
Based on public pay-as-you-go pricing. Enterprise discounts and committed-use contracts may differ. Some providers charge separately for certain features. The calculator uses the public price for the provider configuration that most closely matches Soniox.
Speech infrastructure for massive scale
Build on one API and deploy in your region
Use the same models and API everywhere, with in-region processing to meet latency, data residency, and regulatory requirements.
Available: US, EU, Japan, India
Coming soon: Korea, Australia, Canada, Saudi Arabia, UK, Brazil
Run mission-critical systems with confidence
- 99.9% uptime
Production-hardened infrastructure with monitoring and redundancy. - low-latency streaming
Process speech in real time with low latency for responsive voice applications. - Priority support
Severity-based incident response with direct access to the Soniox team.
"Before Soniox, our international users always had a noticeably different experience. Now accuracy and responsiveness match across all regions…it feels like one system instead of five."
Alon Yair CTO of Onvego
Build products that speak more than one language
Use language mixing wherever speech crosses language boundaries, from multilingual voice agents and customer communication to localization, bilingual learning, narration, and accessible reading experiences.
Multilingual communication
Speak messages that mix languages, foreign names, local addresses, and international terms without splitting the text into separate requests.
Voice agents
Let voice agents speak customer names, addresses, product terms, and multilingual responses naturally when languages change inside a conversation.
Video voiceover and localization
Generate localized voiceovers that preserve foreign names, brands, places, and phrases naturally inside the surrounding language.
Language learning
Build bilingual lessons and explanations where one voice moves naturally between the learner’s language and the language being taught.
Audiobooks and narration
Narrate books and long-form content containing multilingual names, dialogue, places, and borrowed phrases without changing voice identity.
Accessibility and reading
Read mixed-language documents and interfaces aloud while preserving names, foreign terms, and language changes in the source content.
Privacy and compliance, built right in
Never stored, never saved.
Audio stays in memory, everything is processed in real-time.
Built for privacy-critical use cases.
Adhering to leading global security, privacy, and compliance standards.
Trusted where privacy matters most.
Used in industries where speech is sensitive, from healthcare to enterprise.
Frequently asked questions about language-mixing text-to-speech
What is language mixing in text-to-speech?
Can Soniox speak multiple languages in the same sentence?
Does Soniox Text-to-Speech support code-switching?
Do I need to split mixed-language text into separate TTS requests?
Which language should I set in the request?
language field is required. Set it to the main language of the text, or to the language you want to drive the overall delivery. It controls the overall prosody and applies a light accent bias to the utterance.Do I need to change the language setting for foreign names and phrases?
Can Soniox pronounce foreign names and places naturally?
Does the same text-to-speech voice work across languages?
Do cloned voices support multilingual language mixing?
Can language mixing be used with streaming text-to-speech?
Which languages does Soniox Text-to-Speech support?
How much does Soniox Text-to-Speech cost?
How do I start building with language-mixing TTS?
Send your mixed-language text directly to Soniox Text-to-Speech API and set the language field according to the primary or desired delivery language.
Read the language mixing documentation for request examples and implementation details.
Speak every language change naturally
Build multilingual speech with one Text-to-Speech API. Send mixed-language text in one request, keep one consistent voice across 60+ languages, and generate natural speech without splitting text or stitching audio.
Ready to get started?
Create an account instantly, or contact us to design a custom package for your business.
Build with APIDocumentation
Get up and running in minutes and spend your time building, not wrestling with the API.
Explore docsSee what you’ll pay
Pay only for what you use with our flexible pricing. Built to scale with you.
Pricing details