Google has launched Gemini Avatar, a new feature inside Google Flow AI that scans your face and voice to create a digital clone of you, then lets you insert that clone into AI-generated videos featuring any scene or setting you imagine. The feature, announced at Google I/O 2026 in May, is now rolling out to paid subscribers throughout June 2026, and thousands of creators are using it for the first time.

What Exactly Is Gemini Avatar and How Does It Work?

Gemini Avatar is powered by Gemini Omni, Google's newest multimodal AI model that can understand and generate text, images, video, and audio simultaneously. The feature solves one of the hardest problems in AI video creation: keeping the same character looking consistent from scene to scene. Instead of generating a new character for each video, your Avatar stays saved as a reusable asset in your Flow AI project library, meaning you always look like you across every clip and project.

The process is straightforward. First, you scan your face using your phone's front camera, capturing your face's three-dimensional structure from multiple angles. Next, you record your voice by reading a few sentences aloud, allowing Gemini Omni to capture your natural tone, rhythm, and vocal characteristics. Once processed, your Avatar is stored in Flow AI like any other creative asset. When you generate a video, you simply type @me in your prompt, and Veo 3.1, Google's video generation model, creates the final clip with your Avatar as the main character, complete with full motion and synchronized audio.

Every video generated with your Avatar includes an invisible SynthID watermark from Google DeepMind, marking the content as AI-generated for transparency and safety purposes.

What Do You Need to Get Started With Gemini Avatar?

Access to Gemini Avatar requires specific hardware, software, and subscription requirements. Here's what you'll need:

• Subscription Level: A paid Google AI Plus subscription ($19.99 per month, or approximately 1,400 Pakistani rupees per month). The Avatar feature is not available on the free tier of Google's AI services.
• Platform Access: Either the Google Flow AI mobile app (available on the App Store or Google Play) or the desktop browser version at flow.google. The Avatar setup process works on mobile where you have a front camera for the face scan, and the resulting Avatar asset can be used on both mobile and desktop.
• Proper Lighting: Good lighting during the face scan makes a significant difference in Avatar quality. A well-lit room with natural daylight or a lamp positioned directly in front of your face produces a substantially better Avatar than a dark environment or a room where light comes from behind you.
• Quiet Environment: A quiet space for voice recording is essential, since background noise during voice capture affects the quality of your voice clone. Air conditioning, traffic, or background television can all degrade the voice model.

How to Set Up Your Gemini Avatar in Google Flow AI

The setup process takes roughly 10 to 20 minutes from start to finish, though processing time can extend to 10 minutes after you complete the initial scans. Here are the exact steps:

• Sign In and Navigate: Open flow.google in your browser or launch the Google Flow AI mobile app, then sign in with your personal Google account. Personal Gmail accounts work best; Google Workspace accounts (school or business) may have restrictions on Flow AI features.
• Locate the Avatar Section: Inside Flow AI, look for the Avatar or Characters section in the project assets panel, the same area where you'd find saved Ingredients and Style references. On mobile, it may appear as a dedicated button; on desktop, check the sidebar or assets panel for an "Add Avatar" or "Create Avatar" option.
• Complete the Face Scan: Tap or click Create Avatar and position your face so it fills most of the camera frame. Follow the on-screen instructions to look directly at the camera, slowly rotate your head left and right, and tilt your chin slightly up and down. The entire face scan takes 30 to 60 seconds. Move slowly and smoothly, as jerky movements can affect scan quality.
• Record Your Voice: After the face scan, you'll be prompted to record your voice. Read the on-screen text at your normal speaking pace and volume without trying to sound different or unusual. The system captures your natural voice, and the more naturally you speak, the better the Avatar's voice will match you. Recording takes about 1 to 2 minutes.
• Wait for Processing: Gemini Omni processes your data to build your Avatar, typically taking 3 to 10 minutes. You'll receive a notification inside the app when your Avatar is ready. You don't need to keep the app open during processing.
• Review and Finalize: When processing is complete, you'll see a preview of your Avatar with a short generated clip showing your Avatar speaking or moving. If you're satisfied with the result, you're ready to use it. If the Avatar doesn't look quite right, you can redo the scan; most people find the first attempt is good enough to work with.

How Does Gemini Avatar Compare to Other AI Assistants?

While Gemini Avatar focuses specifically on video generation with personalized digital clones, Google's broader Gemini AI framework competes with other major AI assistants like Apple's Siri AI. The key difference is architectural: Gemini operates as a cloud-first system with access to Google's vast multimodal model infrastructure, allowing it to handle text, coding, images, and video simultaneously with a context window of over 1 million tokens, meaning it can process roughly 100,000 words at once.

In contrast, Apple's Siri AI uses a hybrid approach with on-device processing for regular requests and Private Cloud Compute for advanced queries, prioritizing data privacy over raw processing power. Gemini's strength lies in its versatility across platforms, live web access through Google Search integration, and ability to generate photorealistic images and comprehensive content. For creators specifically interested in video generation with personalized avatars, Gemini Avatar inside Flow AI offers capabilities that go beyond what traditional voice assistants provide.

What Are the Real-World Implications for Content Creators?

Gemini Avatar removes a major barrier to video content creation: the need to appear on camera. Creators can now generate cinematic videos of themselves standing on a beach at sunset, presenting on a stage, or exploring a fantasy forest, all without recording themselves in person. This democratizes video production for creators who are camera-shy, have limited equipment, or want to generate content at scale.

The feature's integration directly into Flow AI's creative workflow means creators can use their Avatar the same way they use other tools in the platform, such as the Ingredients to Video system and Veo 3.1 video generation. The invisible SynthID watermark ensures transparency about AI-generated content, addressing growing concerns about deepfakes and synthetic media authenticity. As Gemini Avatar continues rolling out through June 2026, the feature is expected to reshape how creators approach video production, particularly for those building personal brands, educational content, or marketing materials.

" }