Providers
Inworld
Inworld text-to-speech with streaming and multilingual voices.
| Prefix | inworld |
| Default model | inworld-tts-1.5-max |
| Env var | INWORLD_API_KEY |
| Official docs | docs.inworld.ai |
Models
| Model | Streaming | Audio Tags | Voice Cloning | Notes |
|---|---|---|---|---|
inworld-tts-1.5-max | Yes | No | No | Flagship Llama TTS 1.5 Max — best balance of quality and speed |
inworld-tts-1.5-mini | Yes | No | No | Llama TTS 1.5 Mini — ultra-fast, most cost-efficient |
Both models support 15 languages: English, Chinese, Japanese, Korean, Russian, Italian, Spanish, Portuguese, French, German, Polish, Dutch, Hindi, Hebrew, and Arabic, with enhanced timestamps.
Usage
import { generateSpeech } from "@speech-sdk/core"
const result = await generateSpeech({
model: "inworld/inworld-tts-1.5-max",
text: "Hello from SpeechSDK!",
voice: "Ashley",
})Built-in voices include Ashley, Dominus, Edward, Hades, Priya, and others — see the Inworld docs for the full list.
Provider Options
await generateSpeech({
model: "inworld/inworld-tts-1.5-max",
text: "Hello!",
voice: "Ashley",
providerOptions: {
temperature: 0.8,
language: "en",
},
})Custom Configuration
import { generateSpeech } from "@speech-sdk/core"
import { createInworld } from "@speech-sdk/core/providers"
const inworld = createInworld({
apiKey: process.env.INWORLD_API_KEY,
})
const result = await generateSpeech({
model: inworld(),
text: "Hello!",
voice: "Ashley",
})