Moviva

VRM avatar with real-time lip sync in place of the ChatGPT voice orb.

Moviva turns ChatGPT Voice into a face-to-face conversation. When you start a voice chat on chatgpt.com, an animated character appears in a video-call window: it speaks with ChatGPT's voice, listens while you talk, and reacts to what is being said. HOW IT WORKS 1. Open chatgpt.com and click the Moviva icon to turn the avatar on. 2. Start ChatGPT Voice as usual. 3. The call window appears. Talk. When the voice chat ends, the window disappears. ChatGPT's own controls (text field, mute, end call) stay visible and usable. Moviva adds no controls of its own to the call. WHAT THE AVATAR DOES • Lip sync in real time with ChatGPT's voice. • Listens when you speak: nods, keeps eye contact, stops gesturing the moment you interrupt. • Gestures that fit the reply: a question, agreement, a list, an explanation and a conclusion each look different. • Emotions on the face: smiles, laughter, surprise, sadness, held for a few seconds like a real reaction. • Tone of voice matters: the same sentence said calmly and excitedly is acted differently. • Natural idle behaviour: breathing, blinking, shifting weight, glancing around. • Notices the page: may glance at a new message or follow your mouse pointer near its face. • A camera that moves gently with the conversation. THE CALL WINDOW • Opens full-screen by default; drag to move it, drag a corner to resize. • Pin it to float above other tabs and apps (Picture-in-Picture), so you can keep talking while you work elsewhere. MOVIVA FREE The full experience with one built-in avatar. No time limits, no session limits, no account. MOVIVA PLUS (one-time purchase, no subscription) • Character library: additional avatars to choose from. • Add your own avatar: bring a compatible VRM model. • Your own call background image. • Character profiles: Natural, Warm, Professional, Calm, Playful, Anime. Each changes how the avatar acts. • Inject Role: one click asks ChatGPT to speak in the chosen character and language. PRIVACY • Audio, voice analysis and reply text are processed locally in your browser. Nothing from your conversations is recorded, stored or sent to us. • No analytics, no ads, no tracking in the extension. • Moviva Free makes no network requests to us at all. Moviva Plus contacts our licensing server only to activate and renew the licence and to download the avatar pack. • The optional emotion model is downloaded once from Hugging Face and then runs offline. • Microphone reactions are optional and can be switched off at any time. REQUIREMENTS • Chrome 116 or newer on desktop. • A ChatGPT account with Voice available on chatgpt.com. • Headphones are recommended, so the avatar does not mistake ChatGPT's voice for yours. GOOD TO KNOW • Moviva works only on chatgpt.com and only during a voice chat. • Local text analysis for gestures understands English, Russian, Ukrainian and Spanish. In other languages the avatar reacts to the voice. Moviva is an independent product. It is not affiliated with, endorsed by or sponsored by OpenAI. ChatGPT is a trademark of OpenAI.