Gradium

Gradium is a text-to-speech plugin. It streams low-latency voices in English, French, German, Spanish, and Portuguese, plus your own clones.

Setup

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

export GRADIUM_API_KEY=<key>

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

from zeroruntime.plugins import GradiumTTS

tts = GradiumTTS()

# Pipeline(tts=tts, ...)

GradiumTTS ships in the Python SDK only. The Node JS SDK does not include this plugin yet.

GradiumTTS requires a zeroruntime release newer than 0.1.9.

Constructor parameters for GradiumTTS.

Parameter	Type	Default	Description
`api_key`	`str`	`None`	Gradium API key. Falls back to the `GRADIUM_API_KEY` environment variable when unset.
`model`	`str`	`"default"`	TTS model: “default” (stable) or “gradium-tts-beta” to follow betas.
`voice_id`	`str`	`"YTpq7expH9539ERJ"`	Gradium voice ID, from the catalog or your own voices.
`temperature`	`float \| None`	`None`	Output variety from 0 to 1.4. Leave unset for Gradium’s default (0.7).
`padding_bonus`	`float \| None`	`None`	Speaking speed from -4 to 4. Negative is faster, positive slower. Leave unset for Gradium’s default (0).
`rewrite_rules`	`str \| None`	`None`	Text rewriting rules, e.g. “en”, or “none” to disable them. Leave unset to use the voice’s language.
`pronunciation_id`	`str \| None`	`None`	Pronunciation dictionary to apply. Leave unset for none.
`base_url`	`str`	`"wss://api.gradium.ai/api/speech/tts"`	WebSocket endpoint of Gradium streaming text-to-speech. Change only for a proxy or a regional endpoint.

SDK	Import	Constructor
Python	`from zeroruntime.plugins import GradiumTTS`	`GradiumTTS()`

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