AI VTuber TTS: Give your AI VTuber a Kitsch Labs voice.

Choose a voice for your character and turn LLM replies into speech. Connect Kitsch Labs TTS to bring character to your AI VTuber’s greetings and conversations.

Product demos

Enter a line for your AI VTuber45/200 · Up to 200 characters

By generating speech, you agree to Kitsch Labs’ Terms of Service and Privacy Policy and confirm that you are at least 18 years old.

LLM replies, in your character’s voice

Every conversation gives your AI VTuber something new to say. Your LLM writes the reply using the character’s personality and context. Kitsch Labs TTS reads it in the voice you choose.

Keep your LLM and character settings, and connect Kitsch Labs where speech is generated. Try real lines, from a first greeting to a conversation with viewers, to find a voice that fits.

Eunha greeting viewers at an outdoor screening

Connect a voice to an open-source AI VTuber

Start with an open-source project such as Open-LLM-VTuber, which brings conversation and an avatar together. Projects with configurable OpenAI-compatible TTS can use an API URL, key, model, and voice to configure the connection. Choose a Kitsch Labs voice and hear your first greeting.

Explore Open-LLM-VTuber

For the AI VTuber you build yourself

Connect the TTS API to your own conversation system. Send the spoken part of an LLM reply with a Voice ID, then play the returned audio through your avatar’s playback flow. Connect lip sync and movement using your avatar tool’s audio input.

Explore the TTS API

Character voices made with professional voice actors

From a cheerful greeting to a calm conversation, compare character voices made with professional voice actors. Listen to samples in the voice catalog and check each voice’s actor information and supported languages.

Browse character voices

How to connect Kitsch Labs TTS to an AI VTuber

1. Prepare your LLM and AI VTuber project

Install an open-source project such as Open-LLM-VTuber or prepare your own conversation system. Connect your LLM and set the character’s personality and speaking style. Follow the installation and LLM instructions for your project version.

2. Get an API key and Voice ID

Create an API key in Kitsch Labs. Choose a character on the home screen and use the copy button at the top right of the character to copy its Voice ID. Check the supported languages and voice usage conditions.

3. Configure OpenAI-compatible TTS

Select OpenAI-compatible TTS and enter your Kitsch Labs Base URL and API key. Set the model to kitsch-voice and Voice to the copied Voice ID. Use the openai_tts configuration example and Speech API docs to match the audio format. Configure the LLM and TTS connections separately.

4. Connect LLM replies to audio playback

In an open-source project, check that conversations call your configured TTS. In your own implementation, pass the LLM reply as input to the Speech API and play the returned audio. Send only spoken dialogue, leaving out stage directions and internal instructions. Keep API keys in server or local configuration.

5. Have your first conversation

Start with a short request such as ‘Introduce yourself to the viewers.’ Listen for your selected voice and adjust name pronunciation, reply length, and avatar playback timing. Check response time and audio order in your conversation environment.

Why choose Kitsch Labs for your AI VTuber’s voice?

A voice that fits the character

Compare voices using greetings and replies your character will actually say.

Read the character voice guide

Made with professional voice actors

Explore how character planning and performance shaped UNLISTED’s four voices.

Read the collaboration story

An API from LLM text to speech

Request audio with reply text and a Voice ID. Find the connection fields in the OpenAI-compatible Speech API docs.

Read the Speech API docs

Frequently asked questions

Which LLMs can I use?

You can build an integration if your system can pass text replies to the TTS API. For open-source projects, check the LLMs and TTS connections supported by your project.

Will any OpenAI-compatible project work immediately?

Check that the project supports OpenAI-compatible TTS and lets you change the Base URL, API key, model, and Voice. Compatibility for LLM requests alone may require a separate TTS integration. Also check your version’s configuration and audio format.

Does Kitsch Labs provide the avatar and lip sync?

Kitsch Labs generates speech from your text. Connect the avatar display, movement, and lip sync through your AI VTuber project or avatar tool.

Does this change my microphone voice in real time?

This is TTS: it generates a character voice from typed text or LLM replies. It works differently from a voice changer that transforms microphone audio.

Can I use Korean and Japanese?

Language support varies by voice. Check languages and samples in the voice catalog, then compare pronunciation and delivery using representative lines in your target language.