Resources
FAQ

Frequently Asked Questions

General

What is Avatarium?

Avatarium is real-time 3D AI avatar infrastructure. Embed a lifelike AI-powered avatar on any website or app with a single line of code. Use cases include customer support, coaching, education, sales demos, and interactive NPCs.

Do I need 3D modeling experience?

No. Avatarium includes a ready-to-use avatar library. For custom avatars, you can import from Ready Player Me or upload GLB files.

What AI models are available?

Standard (all plans): GPT-4o Mini, Claude Haiku 3.5, Gemini 3.1 Flash, Grok 3 Mini

Premium (Premium/Elite plans): GPT-5.4, Claude Sonnet 4.6

All plans also support BYOK – bring your own OpenAI, Anthropic, Google, or xAI API keys.

Plans & Billing

What plans are available?

See avatarium.ai/pricing (opens in a new tab) for current plan details. Plans range from Free (30 min/mo) through Elite ($90/mo, 720 min/mo).

Can I use my own API keys?

Yes. All plans support BYOK. When using your own keys, AI tokens are billed directly by that provider.

What counts as a voice minute?

Minutes are counted when the user is speaking (STT) or the avatar is responding (TTS). Idle time is not counted.

What happens when I hit my limit?

You'll be notified at 70% and 90% of your monthly limit. If you exceed your included minutes, usage stops until your next billing cycle. Upgrade anytime to continue immediately.

Safety & Control

What if my avatar says something incorrect or inappropriate?

You have multiple layers of control:

  1. Knowledge Base – Ground responses in your approved content
  2. Topic Boundaries – Define topics the avatar should refuse
  3. Human Escalation – Auto-transfer to a human agent when needed
  4. Safety Filters – Built-in harmful content protection

See Safety & Content Controls.

Can I review conversations?

Yes. All conversations are logged and available in Dashboard → Analytics → Conversation Transcripts. You can search, review, and export history.

Can the avatar hallucinate?

AI can generate inaccurate information. Minimise this with a well-defined Knowledge Base, clear system prompts, and confidence threshold escalation.

Technical

What browsers are supported?

Chrome 90+, Firefox 90+, Safari 14+, Edge 90+.

Does it work on mobile?

Yes. Avatars are fully responsive on iOS and Android browsers. Voice input requires microphone permission.

What's the latency?

Average response latency under 200ms globally, including AI processing and TTS.

What is the avatar embed URL format?

https://avatarium.ai/a/{avatarShortId}

Short IDs are 11-character YouTube-style strings (e.g. 8yb8p7oj3sK).

Can I self-host?

Enterprise plans include self-hosting options. Contact sales@avatarium.ai.

Privacy & Security

Where is my data stored?

Data is stored in secure US data centers, encrypted at rest and in transit.

Are conversations private?

Yes. Conversations are encrypted in transit and at rest. Conversation data is never used to train AI models.

Support

Contact Support →

ChangelogTroubleshooting