This section provides the full phoneme tables for each language supported by Capacity Private Cloud Neural and LLM Text-to-Speech. Each table lists every phoneme with its Unicode code point(s), decimal value(s), IPA symbol, X-SAMPA representation, viseme, and an example word, with consonants and vowels listed separately.
Use these phonemes in SSML with the <phoneme> element (with alphabet="ipa" or alphabet="x-sampa") to override the default pronunciation of a word. For the list of voices available in each language, see Neural TTS Languages.
Supplementary English phonemes. Several non-English phoneme sets (for example Romanian, Italian, and the Nordic languages) also include a number of English phonemes. These are provided so that English words and loanwords embedded in the target-language text are pronounced correctly, and are marked with * in each article.
English
French & German
Iberian
Celtic
Italian & Portuguese
Nordic
Central & Eastern European
East Asian
