Skip to content

Put a real-time avatar in the device you build

Essence 2 and Expression 2 ship as SDKs that render on the device's own processor: Android (arm64, from maven.bithuman.ai), iPhone, iPad and Mac (Swift package), and Linux or macOS (CLI and Python).
  • Network

    Devices check in when a session starts and keep rendering through a network drop of up to 5 minutes. On Android, after the one-time model download, the only network traffic is usage reporting. Fully disconnected operation is available on Linux under an offline license.

    Essence 1 runs fully offline today on Linux (x86_64 and ARM64) and on macOS with Apple silicon. Essence 2 and Expression 2 run fully offline on Linux x86_64 (bitHuman 2.11.17 or later). On Linux, the Python package and the bitHuman CLI (2.8.4 or later) both run offline packs; on a Mac, the Python package. Expression 1 runs in the bitHuman cloud only.

  • Credentials

    Each device authenticates with an API secret when a session starts. Use a separate secret for each device fleet, and rotate it if a device is lost.

    API secrets (docs)

Where it renders

On the device, on your own servers, in the bitHuman cloud, or fully offline. Compare the four modes

Tell us where it needs to run