Normalize Text
Session Operations
Normalize Text
Apply text normalization rules to prepare text for TTS.
What It Does
Text normalization converts text into a format optimized for speech synthesis:
- Expands abbreviations (e.g., “Dr.” → “Doctor”)
- Converts numbers to words (e.g., “123” → “one hundred twenty-three”)
- Handles special characters and symbols
- Applies locale-specific pronunciation rules
Request Parameters
| Parameter | Required | Description |
|---|---|---|
text | Yes | The text to normalize |
locale | No | Target locale for normalization (e.g., ko, en, ja) |
Example
Select “NormalizeTextRequest” from the Examples dropdown to see the request format.
POST
Normalize Text
Authorizations
Path Parameters
Body
Raw text to normalize. The config's regex rules are applied in order (lowest first) to rewrite patterns into spoken-form text (e.g. 20℉ → 화씨 20도).
Minimum string length:
1Locale code (e.g. ko-KR). When set, locale-agnostic rules plus rules matching this locale run; when omitted, only locale-agnostic rules run.

