Our website combines advanced technologies to deliver a seamless and interactive experience. Users start by entering a prompt, which is sent to OpenAI's ChatGPT API. The API generates a meaningful response, which is then passed to Eleven Labs, a high-quality text-to-speech (TTS) platform, to convert the text into lifelike audio. This audio file is processed using Rhubarb, a tool that analyzes the speech and extracts phoneme data (speech sounds). The phoneme data serves as the foundation for synchronizing the character's lip movements with the audio.

Using Three.js, a 3D character is rendered on the website, and the extracted phoneme data animates the character's lips for perfect synchronization with the audio. Additional animations, expressions, and movements bring the character to life, creating a dynamic and engaging experience. Built on Vite for fast and efficient performance, the website allows users to see their inputs transformed into sound and animation in real time, blending AI, sound processing, and 3D rendering into an innovative, interactive platform.