SSML and language detection

Voice language detection

When voice_lang is null, the system automatically detects the voice language from message content using the patrickschur/language-detection library.

Supported languages: English (EN), German (DE), French (FR), Italian (IT), Spanish (ES), Polish (PL), Dutch (NL), Romanian (RO), Portuguese (PT), Czech (CS), Hungarian (HU), Swedish (SV), Danish (DA), Finnish (FI), Slovak (SK), Croatian (HR), Turkish (TR), Russian (RU), Bulgarian (BG), Ukrainian (UK).

SSML Support

The API now supports a subset of SSML (Speech Synthesis Markup Language) for advanced voice message formatting.

Supported Features

Available Tags

• <break> - Insert pauses in speech
• <say-as> - Control pronunciation of text

Tag Specifications

<break> Tag

• Purpose: Add pauses between words or sentences
• Attributes: time (e.g., “1s”, “500ms”)
• Pricing: <0.5ms = 1 char, <1s = 2 chars, <1,5s = 3 chars, etc.

<say-as> Tag

• Purpose: Control how text is interpreted and spoken
• Supported Attributes:
  • interpret-as="characters" - Spell out text character by character
• Note: Currently only “characters” value is supported

Usage

Wrap your message content with the <speak> root element:

<speak>
  Your verification code is <say-as interpret-as="characters">A1B2C3</say-as>.
  <break time="1s"/>
  Please enter it now.
</speak>