Lotus Overview
An LLM-enabled Live2D chat companion.
Bring Your Own Model
Select your own models for Chat, Voice, and the Avatar
Run Locally or Connect to Your Favorite AI Provider
You decide! Choose to run a model on your device for privacy and control
or connect to a model provider of your choice.
Upload Your Own Live2D Model
Bring your own chat avatar to life! Upload your custom Live2D model and make
your AI truly unique.
Endless Possibilities
Generate stories, brainstorm ideas, or just have fun chatting with your virtual
AI companion.
How It Works
Lotus is a Windows 64-bit application that connects a Live2D model to a generative AI model, providing a friendly and intelligent avatar to converse with. Unlike most AI chat apps, Lotus allows you to run GGUF LLM models locally, provided you have the hardware to be able to run them.
Lotus focuses on three key aspects to provide this experience: the Avatar, the Chat Model, and the Voice Model. Each of these aspects are customizable.
Avatar
The Avatar is the Live2D model you interact with. You can use any of the built-in models that the Lotus app provides, or you may use your own.
Chat Model
The Chat Model is the backing Large Language Model (LLM) that acts as the brain to the Avatar. You can use popular LLM providers such as ChatGPT or Claude by providing your own API keys, or you can download and run a GGUF model locally, keeping your conversations private on your computer.
Out of the box, Lotus ships with a 1.1B TinyLlama model as the default Chat Model, though it is recommended you download and try a more powerful model to run on your own.
Voice Model
The Voice Model is what gives your Avatar a voice. By default, Lotus uses Piper to synthesize speech, which come with a number of different voices to choose from.
Alternatively, you can provide API keys for ElevenLabs or OpenAI's Text-to-Speech services.