Realistic Lip-Sync 3D Avatar Development in Three.js

We are seeking a Three.js developer to create a realistic web-based 3D avatar with advanced lip-sync functionality. The objective is to build a character that can speak any given text input naturally, with fluent, lifelike mouth movements, and animation. The avatar should look and behave like a real person while speaking, avoiding stiffness or robotic behavior. The developer should integrate a text-to-speech solution like Azure, Amazon Polly, or Google Cloud, and accurately synchronize viseme or phoneme data with the avatar’s morph targets or blend shapes for precise lip movement. Ensuring seamless transitions and realistic timing in the mouth animation is essential to create natural speech flow. The project should use Three.js or React Three Fiber, and a suitable character model can be provided or recommended. The avatar must demonstrate real-time synchronization of mouth, jaw, and facial movements with speech, and include subtle touches such as blinking, breathing, and slight head or eye movement to enhance realism. The code should be clean, modular, and optimized for the web. We require a working demo, the complete source code, and documentation for setup. Experience in 3D animation, morph targets, and speech synthesis APIs is required.

Budget range:
Very small project ($250 – 750 USD)

About the client
I have a task