Ania — frequently asked questions

Pricing, AI creation and wake words

How much does it cost to create a custom talking avatar with Ania?

It depends on the style, resolution and how many states (idle, talking, waving…) you ask for, and the price is always quoted before anything is charged. As a reference, a 2-state avatar (idle + talking) at the default 480p quality is quoted at R$4.20 (about US$0.84): AI cost plus a fixed Ania fee. Ania uses prepaid credits: 1 credit = R$1, shown to non-Brazilian users at a fixed US$0.20 per credit; top-ups are billed in BRL.

How much does a custom wake word cost?

A custom wake word (the phrase that wakes your assistant, e.g. "hey studio") costs US$4.98 (charged as R$24.90). There is a launch discount on a user's first wake word until 2026-12-30. Phrases must have 2–4 words (English also accepts one word with 3+ syllables), no digits, and cannot use other assistants' names such as Alexa or Siri. Wake words are trained for English (en-US) or Brazilian Portuguese (pt-BR).

Can I create an Ania avatar from Claude or another AI assistant?

Yes. Ania has a remote MCP connector (OAuth) that lets an AI assistant plan, quote, create and review an avatar on your Ania account, and search or install marketplace avatars. The assistant always shows the price and cannot confirm a purchase on its own: you approve every spend, and each connection has a monthly spending cap. An npm package, ania-mcp, connects local MCP clients.

What visual styles and movements can an AI-generated Ania avatar have?

Styles: Anime, Comics, Cyberpunk, Realistic, Watercolor, Ghibli-like, Arcane-like, Renaissance, or free text. Movements ("states") include idle, talking, walking, jumping, greeting, farewell, thinking, yes/no, laughing, dancing, surprised and more (18 in total); idle + talking are the default. Each state is a short pre-rendered clip, so playback is light enough to run on ordinary computers.

Is Ania an AI video generator like HeyGen or Synthesia?

No. HeyGen and Synthesia render a presenter video from a script in the cloud. Ania makes a reusable animated character that lip-syncs live to any audio — your microphone, a text-to-speech voice or an AI assistant — on your own device, in a desktop player, a browser extension or a React component. See the comparison page for when each one fits.

Can I sell avatars and wake words I make?

Yes. Published avatars and trained models go through moderation, and creators receive 83% of each sale via Stripe Connect.

The platform

What is ANIA?

ANIA is a platform for creating, sharing, and using animated avatars that react in real-time to your voice. It includes a marketplace for buying and selling avatars, desktop applications (Ania Player and Ania Creators), and a Chrome browser extension for using avatars on any webpage.

How does the voice reactivity work?

ANIA avatars use audio analysis to detect voice input and sync lip movements in real-time. All audio processing happens locally on your device - no audio data is ever sent to our servers. The avatar animation responds instantly to your voice, creating a natural and immersive experience.

What is the ANIA Avatar Chrome Extension?

The ANIA Avatar Chrome Extension allows you to display your animated avatar on any webpage. It includes AI chat integration, Text-to-Speech (TTS) with 5 different providers, Speech-to-Text (STT), and Wake Word detection. You can install it directly from the Chrome Web Store.

Is my voice data private?

Yes, absolutely. All voice and audio processing happens locally on your device. We never transmit, store, or have access to your voice data. Your privacy is our priority - the avatar reacts to your voice without any data leaving your computer.

What TTS (Text-to-Speech) providers are supported?

ANIA supports multiple TTS providers: Browser TTS (free, built-in), TikTok TTS (free), Google Cloud TTS (1 million characters/month free), ElevenLabs (high quality, paid), and Azure (500k characters/month free). You can choose the provider that best fits your needs.

How do I create my own avatar?

You can create custom avatars using Ania Creators, our desktop application available for Windows and Linux. It allows you to design and animate your own unique avatars, which you can then use personally or sell on the marketplace.

Can I sell my avatars on the marketplace?

Yes! Once you create an avatar with Ania Creators, you can upload it to the ANIA marketplace and sell it to other users. Creators receive 83% of each sale directly to their bank account via Stripe Connect. You keep ownership of your creations.

What is the .ania file format?

The .ania format is our custom file format for animated avatars. It contains all the animation frames and metadata needed for real-time voice-reactive playback. You can play .ania files using Ania Player (desktop) or the web player component.

Do I need to pay to use ANIA?

ANIA offers both free and donation-based options. There are free avatars in the marketplace, and the Chrome extension is free to install and use. Premium avatars can be obtained from creators, and you can also create your own avatars at no cost using our tools.

What platforms are supported?

ANIA Player is available for Windows and Linux. The Chrome Extension works on any Chromium-based browser (Chrome, Edge, Brave, etc.). We also provide a React component (ania-avatar-react) for web developers to integrate avatars into their own projects.

How does the Wake Word feature work?

The Wake Word feature allows you to activate your avatar by saying a specific phrase, similar to "Hey Siri" or "OK Google". When the wake word is detected, the avatar becomes active and ready to respond. This processing happens entirely on your device for privacy.

Can I use ANIA for streaming or video calls?

Yes! ANIA avatars are perfect for streaming, video calls, presentations, and content creation. You can use the desktop player as a virtual camera source or overlay, or integrate avatars into your web applications. Many creators use ANIA for gaming streams, educational content, and online meetings.

Who owns the voice I train?

The trained voice belongs to you. You can use it on your avatars, download it and, if you wish, sell it on the marketplace. The platform only hosts and distributes the model according to your choices.

Do I need consent to clone a voice?

Yes. You may only train your own voice or a voice you have express authorization to use. Cloning the voice of a child or of a third party without permission is prohibited. You sign a responsibility statement before you record, and you can request deletion of the data at any time.

What if training fails? Do I lose my coins?

No. If training ends in failure, the charged coins are automatically and fully refunded to your wallet.

How does moderation of published models work?

Every model submitted to the marketplace is reviewed by an administrator before it becomes visible. Models can be rejected or blocked if they break the rules (e.g. missing consent, inappropriate content).

What formats do I receive?

Voices are delivered as .onnx + .onnx.json (Piper). Wake words come as a lightweight .onnx model (~300 KB) plus a manifest. The download is a .zip with a JSON license carrying the purchase details.