xAI

xAI is a text-to-speech plugin. It streams Grok voices and supports inline expression tags like [laugh] and <whisper> in the text.

Setup

Set your xAI API key in the worker environment. Generate a key from the xAI site:

export XAI_API_KEY=<key>

Import the plugin and pass it to the pipeline’s tts slot.

from zeroruntime.plugins import XAITTS

tts = XAITTS()

# Pipeline(tts=tts, ...)

import { XAITTS } from '@zeroruntime/js-sdk/plugins';

const tts = XAITTS();

// Pipeline({ tts, ... })

Constructor parameters for XAITTS.

Parameter	Type	Default	Description
`api_key`	`str`	`None`	xAI API key. Falls back to the `XAI_API_KEY` environment variable when unset.
`voice`	`str`	`"eve"`	Built-in voice (“eve”, “ara”, “rex”, “sal”, “leo”) or a custom voice ID.
`language`	`str`	`"en"`	BCP-47 language code, e.g. “en”, “fr” or “pt-BR”, or “auto” to detect the language from the text.
`codec`	`Literal['pcm', 'mulaw']`	`"pcm"`	Audio encoding of the generated speech: “pcm” (16-bit linear) or “mulaw”.
`sample_rate`	`int`	`24000`	Output sample rate in Hz: 8000, 16000, 22050, 24000, 44100 or 48000.
`optimize_streaming_latency`	`int`	`0`	0 for best quality, or 1 for faster time-to-first-audio with a slight quality tradeoff.
`text_normalization`	`bool`	`False`	Convert numbers, abbreviations and symbols into spoken words before synthesis.
`base_url`	`str`	`"wss://api.x.ai/v1/tts"`	WebSocket URL of the xAI text-to-speech endpoint.
`max_connection_age_sec`	`float`	`300.0`	Seconds after which the streaming connection is reopened, to stay within xAI session limits.

SDK	Import	Constructor
Python	`from zeroruntime.plugins import XAITTS`	`XAITTS()`
Node JS	`import { XAITTS } from '@zeroruntime/js-sdk/plugins'`	`XAITTS()`

Synthesized audio is streamed back to the caller after the LLM produces a response.