Gemini 3.8 Flash TTS turns text into expressive speech, now GA

What you get

Gemini 3.8 Flash TTS and Flash-Lite TTS are generally available as of 2026-09-22, turning text into expressive speech with 150+ voices.

Prerequisites

A Gemini API key in GEMINI_API_KEY
The @google/genai package (npm install @google/genai)

Steps

Choose a model: gemini-3.8-flash-tts for studio-grade, expressive voices, or gemini-3.8-flash-lite-tts for fast, cheaper speech (it replaces gemini-3.1-flash-tts-preview). The same release added a Voices endpoint (/v1beta/voices), voice design and voice replication.
Create speak.mjs. The style annotation directs the delivery and voice picks a prebuilt voice:

javascript

import * as fs from 'node:fs';
import {GoogleGenAI} from '@google/genai';

async function main() {
   const client = new GoogleGenAI({});

   const interaction = await client.interactions.create({
      model: 'gemini-3.8-flash-tts',
      input: [{
         type: 'user_input',
         content: [{
            type: 'text',
            text: 'Have a wonderful day!',
            annotations: [{
               type: 'speech_metadata',
               style: 'cheerful and friendly',
            }],
         }],
      }],
      response_format: { type: 'audio' },
      generation_config: {
         speech_config: [
            { voice: 'Kore' },
         ],
      },
   });

   const audioBuffer = Buffer.from(interaction.output_audio.data, 'base64');
   fs.writeFileSync('out.wav', audioBuffer);
}

await main();

Test it

Run node speak.mjs, then play out.wav. You should hear "Have a wonderful day!" in a cheerful voice.

What you'll need

What you get

Prerequisites

Steps

Test it