Speech Generator

SpeechGenerator converts text to speech, producing an AudioClip. Optionally pairs with a TextToSpeechPlayer to stream audio in real time as chunks arrive.

Component Menu

AI Dev Kit / AI Generators / Speech Generator

Inspector Fields

Field	Description
Settings	SpeechGenerationSettings asset — model and voice
Player	(Optional) TextToSpeechPlayer for real-time streaming playback
Save Outputs	Persist generated audio to the output folder

Events

Event	Signature	Description
onOutput	UnityEvent<AudioClip>	Fires when the complete audio is ready.
onStream	UnityStreamEvent<AudioClip>	Fires for each audio chunk during streaming.
onStatusChanged	UnityEvent<ResponseStatus>	Fires on status change.

Methods

Method	Description
Generate(string text)	Fire-and-forget TTS generation.
GenerateAsync(string text)	Awaitable version. Returns Generated<IAudioAsset>.
Configure(SpeechGenerationSettings)	Override inspector settings at runtime.

Streaming Playback

When a TextToSpeechPlayer is assigned, audio is streamed to the player chunk by chunk. The player stitches audio buffers in real time so playback starts immediately without waiting for the entire clip.

Code Example

using Glitch9.AIDevKit.Generators;
using UnityEngine;

public class TTSExample : MonoBehaviour
{
    [SerializeField] private SpeechGenerator speechGenerator;

    public void Speak(string text)
    {
        speechGenerator.Generate(text);
    }
}