Embed AI avatar conversations in any website. Real-time video, speech recognition, dynamic context injection, and rich visual content (charts, tables, code, diagrams) — all from a few lines of JavaScript.
Both SDKs connect to the same Kaltura AI Avatar backend. Choose based on how much control you need:
Direct Socket.IO + WebRTC connection. You own the video element. Built-in GenUI rendering (charts, tables, code, diagrams). Extensible plugin system. ~100KB + Socket.IO.
Documentation · Live DemoSandboxed iframe embed. Zero dependencies, ~6KB minified. Automatic UI handling. Ideal for quick integrations or when you need browser isolation.
Documentation · Basic Demo<div id="avatar" style="width:800px;height:600px;background:#000"></div>
<script src="https://cdn.socket.io/4.7.2/socket.io.min.js"></script>
<script src="https://cdn.jsdelivr.net/gh/kaltura/conversational-avatar-embed-sdk@latest/sdk-socket/dist/kaltura-avatar-sdk.js"></script>
<script>
const sdk = new KalturaAvatarSDK({
clientId: 'YOUR_CLIENT_ID',
flowId: 'YOUR_FLOW_ID',
container: '#avatar'
});
sdk.on('avatar-speech', ({ text }) => console.log('Avatar:', text));
sdk.connect();
</script>
This repo works as an instant skill for Claude, ChatGPT, Copilot, and any AI coding agent. Point your agent at this repo and tell it what to build:
The agent generates both the app code (HTML + JS) and the Knowledge Base prompt (paste into Kaltura Studio). See AGENTS.md for the full reference.