Download
Amica is an open source interface for interacting with fully animated 3D characters that combine voice chat, vision, and an emotion engine into a single experience. It lets you hold natural conversations with AI characters that can see, listen, and speak, while expressing emotional states through facial expressions and body language. Users can import VRM character models, adjust their appearance, tune the voice to match the character, and define behavior using different large language models and TTS backends. Under the hood, Amica leverages modern web and desktop technologies: three.js and three-vrm for 3D rendering, Transformers.js for running models in the browser, Whisper and Silero VAD for speech recognition and voice-activity detection, and a variety of LLM backends such as llama.cpp servers, ChatGPT-compatible APIs, Ollama, KoboldCpp, and others. It also integrates multiple text-to-speech providers, including ElevenLabs, OpenAI, Coqui, RVC, and AllTalkTTS.
Features
- Highly customizable 3D AI characters with voice, vision, and an emotion engine for expressive interaction
- VRM character import and editing to match the character’s look and personality
- Pluggable LLM backends including llama.cpp, ChatGPT-compatible APIs, Ollama, KoboldCpp, and more
- Multiple TTS providers supported, from cloud services like ElevenLabs and OpenAI to local engines like Coqui and RVC
- Uses Whisper and Silero VAD for robust speech recognition and voice-activity detection in real-time conversations
- Cross-platform deployment via Tauri with desktop builds plus a browser-based experience for mobile, tablet, and desktop
Project Samples
