When using Rime’s models, most text will be correctly pronounced out-of-the box just by writing naturally. Rime’s models understand pronunciation based on context.

However, we do recommend normalizing your text to get consistent and accurate results.

Text normalization is a crucial step in text-to-speech (TTS) systems, where written text is converted into a form that can be accurately and naturally spoken. Normalization ensures the text is clear, comprehensible, and contextually accurate for speech synthesis. Rime also allows you to adjust the pronunciation of questions, pauses, and more by altering the punctuation.

At Rime we have a unique set of guidelines to help you get the best output. To see common use cases, such as numbers, dates, and abbreviations, check the pages in the Formatting Text folder in the menu.