bitHuman makes it possible for any AI to have a face. Real-time lip-synced avatars that work everywhere — from edge devices to cloud, from Raspberry Pi to iPhone.
AI is becoming conversational, but conversations are more than words. People communicate through facial expressions, eye contact, and presence. We believe every AI deserves a face — one that responds in real time, runs on hardware you already own, and feels natural enough to hold a conversation with.
Our platform gives developers and creators the tools to add a visual, animated presence to any AI system. Whether you are building a customer support agent, an educational tutor, a museum kiosk, or a personal companion — bitHuman handles the face so you can focus on the brain.
CPU-based real-time avatar. Runs on anything from a Raspberry Pi to a cloud VM. Optimized for low-latency lip sync with minimal compute.
GPU-powered avatar with full facial expression. Rich, high-fidelity animation driven by our proprietary neural rendering pipeline.
The bithuman Python package on PyPI. Inference engine, session management, and LiveKit integration for server-side deployments.
Native Apple SDK distributed via SwiftPM. On-device audio-only voice pipeline for iOS and macOS apps — no cloud required.
REST API to create agents, start sessions, and manage avatars without running your own infrastructure. Backed by GPU and CPU pools.
Web dashboard for creating agents, managing billing, embedding chat widgets, and monitoring usage — all without writing code.
We believe developer tools should be open and accessible. Our examples, documentation, and integration guides are open source. The Python SDK is distributed via PyPI and the Swift SDK via SwiftPM — standard package managers that developers already use.
We publish quickstart examples for Python, Swift, and web embeddings so you can go from zero to a working avatar in minutes, not days.