About Protoface
Protoface adds high-quality realtime avatars to your AI app or agent, with a drop-in integration for your existing stack. It converts your agent’s audio into realtime face video, so it works across languages with both speech-to-speech realtime models and STT/LLM/TTS pipelines. Protoface’s next-generation models need only a single input image and spin up custom avatars in seconds. The majority of existing low-cost realtime avatar approaches rely on Gaussian-based models, which are much slower to generate custom avatars and lower in fidelity. Protoface solves this problem.Start here
Protoface works with any stack through the REST API, and integrates with many voice AI platforms, SDKs, and frameworks. The Quickstart walks through a complete, runnable agent in a few minutes, and More Integrations lists every supported platform.Quickstart
Run a sample Protoface avatar agent end to end.
LiveKit Agents
Configure the plugin and session lifecycle.
Pipecat
Render Pipecat assistant audio as avatar video.
More Integrations
Quickstarts for Vapi, Agora, VideoSDK, and more.
Reference
Use these pages when you need endpoint details, error handling, limits, or usage data.API Reference
Endpoint schemas generated from OpenAPI.
Errors
Stable error codes and retry behavior.
Quotas
Session caps, usage rounding, and credit behavior.
Usage
Read account-level session and billing totals.