Runnable examples showing how to embed a 3D AI avatar into your app in minutes.
Demo avatar included — no account needed to see it in action. Voice and AI are disabled on the demo avatar. Load your own avatar ID to enable full conversation.
Sign up free at dashboard.avatarium.ai → create your avatar → copy your Avatar ID → paste it into any example below.
| Example | Framework | Command |
|---|---|---|
| embed/ | Widget embed (2 lines of HTML) | open embed/index.html |
| js/ | JS SDK (no framework) | cd js && npm run dev |
| react/ | React + Vite | cd react && npm install && npm run dev |
| nextjs/ | Next.js | cd nextjs && npm install && npm run dev |
| vue/ | Vue 3 + Vite | cd vue && npm install && npm run dev |
| react-native/ | Expo | cd react-native && npx expo start |
The demo avatar is silent by design. To unlock live voice conversation:
- Sign up at dashboard.avatarium.ai
- Create an avatar and configure:
- AI provider — OpenAI, Anthropic, Google, or xAI (generates responses)
- TTS provider — choose your voice quality tier:
- Basic voice
- Standard voice
- ElevenLabs — Premium
- Copy your Avatar ID from the dashboard
- Paste it into the "Your Avatar ID" field and hit Load
Your avatar will load with full voice conversation enabled. The microphone button activates automatically.
| Package | Install | For |
|---|---|---|
@avatarium/react |
npm i @avatarium/react |
React, Next.js |
@avatarium/react-native |
npm i @avatarium/react-native |
React Native / Expo |
@avatarium/js |
npm i @avatarium/js |
Vue, Vanilla JS, any framework |
- Embedding a 3D avatar in a 16:9 panel
- Loading any avatar by ID at runtime
- Sending text to speak via the SDK (
avatar.speak()) - Controlling playback (stop, volume)
- Reacting to avatar state changes (
ready,speaking,idle) - Demo mode — silent avatar with CTA to create your own
- Dashboard — dashboard.avatarium.ai
- Docs — docs.avatarium.ai
- Website — avatarium.ai