Accented vowels
Radiance · Spanish
Convert Spanish text into natural AI speech for voice agents, customer conversations, narration, applications, media, learning, IVR, and real-time conversational AI. Use Shunya Radiance to generate spoken Spanish from written text with expressive delivery, streaming and batch synthesis, configurable speaking speed, and production-ready audio formats.
Hola. Tu pedido está en camino y llegará mañana por la mañana.
Conversational delivery · RadianceSpanish TTS
Spanish speech synthesis needs to do more than read characters aloud. Names, numbers, dates, abbreviations, punctuation, questions, English terminology, and sentence context all affect how written Spanish should sound when spoken. Radiance Universal speaks Spanish as part of Shunya's world-language speech synthesis.
11
Expression styles across Radiance
<350ms
First-audio streaming latency
7
Supported audio output formats
Streaming + Batch
Two synthesis workflows for real-time and pre-generated speech
Spanish is part of global TTS: Shunya's named 46-voice catalogue covers Indic languages and English. Spanish uses Radiance Universal, with the universal voice Meera or a voice you clone.
| Capability | Radiance · Spanish Text to Speech |
|---|---|
| Spanish Text Input | ✓ |
| Streaming Synthesis | ✓ |
| Batch Synthesis | ✓ |
| Expression Control | ✓ |
| Speed Control | ✓ |
| Multiple Audio Formats | ✓ |
| Web Audio | ✓ |
| Telephony Audio | ✓ |
| API Integration | ✓ |
| Voice Agent Integration | ✓ |
| Cross-lingual Speech Platform | ✓ |
Try it yourself
Don't judge Spanish TTS from one prepared sentence. Try a customer response, address, product name,number, appointment reminder, notification, narration, or conversational response.
Spanish text
Spanish uses the Latin alphabet, but written Spanish contains language-specific characters,accent marks, and punctuation that matter to pronunciation and meaning.
Accented vowels
Ñ
Ü
Question marks
Exclamation marks
¿Cuándo llegará mi pedido?
¡Gracias por comunicarte con nosotros!
El número de confirmación es 48217.
The complete sentence matters more than simply reading individual characters.
Pronunciation needs context
Spanish pronunciation may appear relatively consistent compared with some languages, butproduction TTS still needs to handle much more than ordinary dictionary words.
Customer names, people, companies, and places.
48217, €49.99, 10:30.
15 de septiembre de 2026.
API, CRM, SMS, URL.
Brand names and technical vocabulary may not originate in Spanish.
English terminology frequently appears inside Spanish business and technology content.
A useful Spanish TTS system needs to generate a complete spoken sentence rather than pronounce isolated tokens independently.
Expressive Spanish speech
A voice agent, audiobook, announcement, and customer notification have different delivery requirements. Radiance provides 11 expression styles.
Clean general-purpose speech.
Natural everyday delivery for voice agents and customer interactions.
Long-form storytelling and narration.
Structured informational delivery.
Positive, upbeat speech.
More subdued delivery.
More intense speech.
Anxious or tense delivery.
Reactive and exclamatory speech.
Same text: Tu solicitud ha sido aprobada. - Neutral, Conversational, or Enthusiastic. The words stay the same. The delivery changes.
Spanish + English text
Spanish-speaking users regularly encounter English terminology, especially in technology,software, e-commerce, banking, travel, business, and customer support.
Tu payment se procesó correctamente.
Puedes actualizar tu subscription desde la app.
El API está funcionando, pero necesitamos revisar el login flow.
Tu refund ya fue aprobado.
Applications should expect Spanish text to contain product names, brand names, English technical terms, acronyms, URLs, and numbers.
The application should not need to create a separate audio file every time an English technical term appears.
Real-time Spanish TTS
Fixed announcements can be generated in advance. Conversational AI cannot. Radiance provides streamingsynthesis for applications where speech needs to begin before the complete interaction is over.
<350ms
First-audio streaming latency
Don't wait for the whole paragraph - streaming allows speech playback to begin as audio becomes available.
Streaming
Use streaming when low time-to-first-audio matters.
Best for:
Batch
Generate a full audio output from finished Spanish text.
Best for:
| Requirement | Streaming | Batch |
|---|---|---|
| Voice Agents | ✓ | - |
| Live Conversation | ✓ | - |
| Complete Audio File | - | ✓ |
| Narration | ✓ | ✓ |
| WebSocket | ✓ | - |
| HTTP API | - | ✓ |
Control the audio
The same voice output may need to work in a web application, phone call, media workflow, or voice agent. Radiance supports seven audio formats.
| Format | Common use |
|---|---|
| PCM | Low-latency streaming |
| WAV | Editing and high-quality audio |
| MP3 | Web and general applications |
| OGG Opus | Streaming and web audio |
| FLAC | Lossless workflows |
| μ-law | Telephony |
| A-law | Telephony |
Speed control
Adjust speaking speed across instructional, conversational, and content workflows.
Silence control
Trim unnecessary silence around generated audio where required.
One TTS platform. Different output requirements.
Spanish voice availability
Spanish requests use Radiance Universal, Shunya's text-to-speech model for world languages.Its universal voice, Meera, speaks Spanish and every other supported language.For a brand or custom voice, clone one from a reference recording.
Meera
The universal voice for Spanish and every Radiance Universal language
Model
Set the model value to radiance-universal and the language value to es.
Built-in voice
The universal voice. Meera speaks Spanish and every other language Radiance Universal supports, so one voice can cover a multilingual product.
Custom voice
Clone a brand or custom voice from a reference recording and its transcript, then use it for Spanish speech.
Choose a voice
Pass Meera as the voice value, or send a reference recording to speak in a cloned voice. The same choice works in batch and streaming requests.
Batch and streaming: Spanish synthesis works in HTTP batch and WebSocket streaming requests, with Radiance expression and audio controls.
Cross-lingual voice experiences
Radiance is designed as a multilingual speech-synthesis platform. An application may begin in English. Another customer may need Spanish. A third may speak Hindi. The speech layer should not require a completely separate architecture for every language.
One platform. Multiple language experiences. A cloned voice keeps the same identity as the language changes, so one speaker can carry Spanish and your other languages.
Spanish voice agents
A voice agent needs to listen before it can speak.
Quiero saber dónde está mi pedido.
Converts speech into text.
Understands the request and generates a Spanish response.
Generates spoken Spanish.
Tu pedido está en camino y llegará mañana.
STT → LLM → TTS
Use for:
Real-world use cases
Generate dynamic Spanish responses from conversational AI.
Provide automated Spanish-language support experiences.
Generate menus, notifications, confirmations, and dynamic phone responses.
Speak order updates, product information, delivery messages, and support responses.
Generate spoken Spanish information for guests, travellers, and customer-service applications.
Create Spanish audio for lessons, instructions, exercises, and educational content.
Generate Spanish narration for digital content and production workflows.
Provide spoken output for Spanish written interfaces.
Create voice reminders, payment updates, delivery alerts, and appointment messages.
Add Spanish speech generation to existing applications without building separate audio infrastructure.
Developers
Radiance provides HTTP batch and WebSocket streaming synthesis. Batch endpoint: POST https://ttsv2.shunyalabs.ai/v1/audio/speech
{
"model": "radiance-universal",
"language": "es",
"input": "<Conversational> Hola. ¿En qué puedo ayudarte hoy?",
"voice": "Meera",
"response_format": "mp3"
}Basic workflow
Real-time workflow
Configure
Spanish requests use Radiance Universal: set model to radiance-universal, language to es, and voice to Meera, the universal voice, or clone a custom voice from a reference recording.
Built for production
Generating one Spanish audio clip is simple. Production TTS systems need low-latency streaming, batch generation,styles, audio controls, APIs, telephony formats, multilingual expansion, and enterprise deployment.
| Requirement | Radiance |
|---|---|
| Spanish Text to Speech | ✓ |
| World-language Coverage | ✓ |
| Spanish Text Input | ✓ |
| Batch Generation | ✓ |
| Streaming Generation | ✓ |
| First Audio | <350ms |
| Expression Styles | 11 |
| Speed Control | 0.25× - 4.0× |
| Audio Formats | 7 |
| MP3 | ✓ |
| WAV | ✓ |
| PCM | ✓ |
| OGG Opus | ✓ |
| FLAC | ✓ |
| μ-law / A-law | ✓ |
| API Integration | ✓ |
| LLM → TTS Workflow | ✓ |
| Voice Agent Integration | ✓ |
| Enterprise Deployment | ✓ |
Questions
Spanish Text to Speech converts written Spanish into spoken audio using speech-synthesis technology.
Yes. Spanish is synthesized by Radiance Universal, Shunya's text-to-speech model for world languages.
No. Spanish is synthesized by Radiance Universal. The named Indic catalogue is a different language and voice set.
Spanish TTS accepts written Spanish text, which commonly includes characters such as á, é, í, ó, ú, ñ and ü, as well as Spanish punctuation.
Spanish uses one built-in voice, Meera, the universal female voice in Radiance Universal, which speaks every language the model supports. For a male, brand, or custom voice, you can clone one from a reference recording and its transcript.
Yes. Radiance supports WebSocket streaming.
Streaming TTS delivers first audio in under 350 ms. Actual latency can vary with infrastructure, configuration, text, network conditions, and voice type.
Yes. Radiance supports 11 expression styles, including Neutral, Conversational, Narrative, News, Happy, Sad, Enthusiastic, and others.
Yes. Radiance supports speed control from 0.25× to 4.0×.
Radiance supports PCM, WAV, MP3, OGG Opus, FLAC, μ-law, and A-law.
Yes. μ-law and A-law formats are available for telephony-oriented workflows.
Yes. Real-world Spanish frequently contains brands, acronyms, product terms, and English technology vocabulary. Test your own mixed-language text to evaluate how the specific terms used by your application are pronounced.
Yes. Streaming TTS can provide the speech-generation layer for Spanish conversational AI.
Shunya's global STT model covers international languages, so Spanish belongs in the broader global speech-recognition stack as well.
Parallax, Shunya's translation model, focuses on 55 Indian languages and 2,970 translation pairs. Spanish text to speech is a Radiance feature and does not include translation.
Yes. Radiance provides HTTP batch synthesis and WebSocket streaming workflows.
Get started
Don't judge Spanish TTS from our prepared sentence. Test yours. Try a customer response, product name, appointmentreminder, price, notification, narration, Spanish-English sentence, voice-agent response, or text generateddynamically by your application. Hear how Shunya turns your Spanish text into speech.