use base64::{engine::general_purpose::STANDARD as BASE64, Engine};
use display_error_chain::DisplayErrorChain;
use gemini_rust::prelude::*;
use std::fs;
use std::process::ExitCode;
use tracing::info;
#[tokio::main]
async fn main() -> ExitCode {
tracing_subscriber::fmt()
.with_env_filter(
tracing_subscriber::EnvFilter::builder()
.with_default_directive(tracing::level_filters::LevelFilter::INFO.into())
.from_env_lossy(),
)
.init();
match do_main().await {
Ok(()) => ExitCode::SUCCESS,
Err(e) => {
let error_chain = DisplayErrorChain::new(e.as_ref());
tracing::error!(error.debug = ?e, error.chained = %error_chain, "execution failed");
ExitCode::FAILURE
}
}
}
async fn do_main() -> Result<(), Box<dyn std::error::Error>> {
let api_key =
std::env::var("GEMINI_API_KEY").expect("GEMINI_API_KEY environment variable not set");
let client = Gemini::new(api_key)?;
info!("text-to-speech interaction example starting");
let interaction = client
.create_interaction()
.with_model("gemini-3.1-flash-tts-preview")
.with_text("Say hello in a cheerful voice: 'Hello! Welcome to the world of AI!'")
.with_speech_config(InteractionSpeechConfig {
voice: Some("Charon".to_string()),
language: None,
speaker: None,
})
.execute()
.await?;
info!(response = interaction.output_text(), "text response");
if let Some(InteractionContent::Audio {
data, mime_type, ..
}) = interaction.output_audio()
{
if let Some(ref audio_data) = data {
let audio_bytes = BASE64.decode(audio_data)?;
let ext = match mime_type {
Some(AudioMimeType::Mp3) => "mp3",
Some(AudioMimeType::Wav) => "wav",
_ => "pcm",
};
let filename = format!("generated_speech.{ext}");
fs::write(&filename, audio_bytes)?;
info!(filename = filename, "audio saved");
}
}
info!("text-to-speech example completed");
Ok(())
}