Skip to main content
Talking Avatar, by InterviewFlowAI, gives AI voice agents a talking face that’s driven by audio. It works with any voice: OpenAI, ElevenLabs, Azure, Fish Audio, Cartesia, your own text-to-speech, or a speech-to-speech model. The avatar follows the audio, so you don’t need text, phoneme timings or viseme events. There are three ways to use it.

Realistic Avatars

For developers. $0.02 a minute. Photorealistic avatars that speak your agent’s voice, streamed to your page over WebRTC or into your LiveKit room.

Voice Agents

No code. $0.06 a minute. Write a prompt, pick a face and a voice, then share a link or embed the agent on your website. The AI and the voice are included.

Open Source

Free, MIT licensed. A 3D talking avatar that runs your own .glb model in the browser. No server, no account and no API key.
New accounts get $1 of free credits: 50 minutes of Realistic Avatars, or about 16 minutes of Voice Agent calls. You can also talk to a realistic avatar for two minutes without signing up.

​
What you can build

  • AI interviewers that run structured screening calls on video.
  • Customer support and sales agents that greet visitors on your website.
  • Tutors and language-practice partners that explain, listen and give hints.
  • Receptionists that answer questions and take appointment requests.
  • Presenters for product tours and onboarding.

​
Start here

Quickstart

Create an account and have a talking avatar running in a few minutes.

Which product should I use?

Compare Realistic Avatars, Voice Agents and Open Source side by side.

Pricing

Per-second billing from prepaid credits. No subscription and no minimum.

SDK reference

Every class, option, method and event in @interviewflowai/talking-avatar.