README.md

Pronunciation Dictionaries

Control how specific words are spoken using custom pronunciation dictionaries. Useful for names, acronyms, technical terms, and brand names.

Features

• Create a pronunciation dictionary with custom word–pronunciation pairs
• Use the dictionary when generating speech
• List, update, and delete dictionaries
• Compare output with and without custom pronunciation

Requirements

Base dependencies are installed via the root requirements.txt. See the main README for setup. Add SMALLEST_API_KEY to your .env.

Usage

Python

Create a dictionary, synthesize speech with it, then clean up:

uv run python/pronunciation.py

JavaScript

node javascript/pronunciation.js

How It Works

1. Create a pronunciation dictionary with word–pronunciation pairs
2. Synthesize speech, passing the dictionary ID in pronunciation_dicts
3. The TTS engine uses your custom pronunciations instead of defaults
4. Clean up by deleting the dictionary when no longer needed

Example Pairs

Word	Pronunciation	Use Case
API	ay pee eye	Acronyms
GIF	jiff	Contested pronunciations
Kubernetes	koo-ber-net-eez	Technical terms
Diya	dee-yah	Names

API Reference

• Pronunciation Dicts API

Next Steps

• Getting Started — Basic speech synthesis
• Voices — Choose the right voice for your use case