Gradium

This repo contains a Rust client for the Gradium Voice AI API.

cargo run -r --example tts -- \
    --text "Hello, this is a test of the gradium text-to-speech system. Please ensure that you follow the signs." \ 
    --out-file ~/tmp/out.wav \

lib.rs:

Gradium Rust Client

A Rust client library for the Gradium API, providing text-to-speech (TTS) capabilities via WebSocket connections.

Quick Start

use gradium::{Client, protocol::tts::Setup, protocol::AudioFormat};

#[tokio::main]
async fn main() -> anyhow::Result<()> {
    let api_key = gradium::api_key_from_env().expect("GRADIUM_API_KEY not set");
    let client = Client::new(&api_key);

    let setup = Setup::new("m86j6D7UZpGzHsNu").with_output_format(AudioFormat::Wav);

    let result = gradium::tts("Hello, world!", setup, &client).await?;
    let filename = "output.wav";
    std::fs::write(filename, result.raw_data())?;
    println!("Generated {} bytes of audio and saved to {}", result.raw_data().len(), filename);
    Ok(())
}

Features

• Simple API: High-level tts() function for one-shot text-to-speech conversion
• Streaming Support: TtsStream for advanced use cases with real-time audio streaming
• Multiple Formats: Support for PCM, WAV, and Opus audio formats
• Async/Await: Built on tokio for efficient async I/O

Environment Configuration

The client expects the GRADIUM_API_KEY environment variable to be set with your API key. You can retrieve it using the [api_key_from_env()] function.