SYNVO Embedded Speech

Embedded Text-to-Speech software with unique polyglot, high-quality, natural-sounding voices for a variety of devices and applications ranging from automotive over consumer electronics to assistive technologies.

  • Unique, sophisticated polyglot capabilities for processing mixed-lingual texts
  • Scalable memory footprint for usage on various embedded architectures
  • For automotive, consumer electronics and assistive technologies
  • High speech quality and accuracy
  • Support for phonetic alphabets: IPA, X-SAMPA, TeleAtlas, Navteq
  • Application specific pronunciation of abbreviations, proper nouns and text patterns


Polyglot Text-to-Speech

Embedded Speech supports as first tts engine automatic language identification and a sophisticated, real polyglot speech generation of mixed-lingual texts for any mixture of 20+ languages.


Provides a more human-like interaction through natural language understanding, pronunciation accuracy, appropriate breaks for longer sentences, and emphasis on the correct words.

Lifelike prosody even for long texts

SYNVO Embedded Speech is optimized for long texts like newspaper articles or stories, as well as for short prompts and questions.

Direct phonetic input

Allows for optimal and seamless read out of off-line phonetic databases such as navigation map data.