Google Vids’ personal avatar feature, announced on Thursday, now lets users upload a selfie and a voice recording to generate a digital clone of themselves for use in workplace videos, a move that pushes the tool well beyond its origins as a presentation builder.
The update is powered by Gemini Omni Flash, Google’s latest media generation model, which the company says delivers improved text rendering, physics, and realism compared to its predecessors. The model handles everything from generating video from a written prompt and reference images, to swapping out backgrounds, correcting lighting, and adding effects to footage recorded on a smartphone.
Google Vids Personal Avatar Feature and What It Actually Does
The personal avatar sits on top of a capability Google had already been building out. By the end of August 2025, Vids had introduced a catalogue of pre-made photorealistic avatars spanning different ages, genders, and ethnicities, each capable of narrating a script with realistic lip-syncing. The new personal avatar is a separate, additional feature: one that maps to a specific user’s own face and voice.
Access is restricted. Personal avatars are limited to users in certain regions who are 18 or older, and each avatar is tied to the account holder’s Google account. Every piece of content generated with the tool is invisibly watermarked via SynthID, the watermarking technology developed by Google DeepMind that embeds markers directly into AI-generated images, audio, text, and video without affecting quality.
Usage limits vary by licence tier. Users with AI Expanded Access add-on licences receive higher usage allowances for Gemini Omni in Vids than those on standard Google Workspace plans.
The broader Gemini Omni integration also introduces step-by-step editing: users can refine a video iteratively rather than discarding the entire project and starting again. Google positions Gemini Omni as a replacement for Veo in the Gemini app, combining the model’s core intelligence with generative media capabilities that include image-to-video conversion and AI-driven video editing.
From Presentation Tool to Video Platform
The updates also arrive alongside an expansion of the avatar roster itself. Google Vids added 12 new cartoon-style avatars in both 2D and 3D formats, plus seven new languages enabling AI-generated audio narration beyond the original supported set.
Taken together, the changes represent a meaningful broadening of what Vids is designed to do. When Google launched it, the pitch was squarely at the enterprise: a tool to help teams produce company updates, onboarding materials, and training content inside Google Workspace without needing a production crew. That framing remains intact, Vids is still a Workspace product, built around business use cases.
But personal avatars and conversational, iterative editing pull it into territory occupied by a different set of rivals. Companies such as HeyGen, Synthesia, Captions, and D-ID have built businesses around exactly this proposition: letting individuals and organisations create presenter-led videos from text, without appearing on camera themselves. Google, with Workspace distribution already embedded across millions of organisations, now has a direct answer to each of those products.
The competitive context also involves at least one notable absence. OpenAI’s Sora, which had allowed users to generate AI videos and had famously permitted people to create clips featuring public figures, including chief executive Sam Altman, has since shut down. Google is arriving at a moment when one of the most discussed tools in the category is no longer available.
The guardrails Google has placed around personal avatars, account binding, age verification, regional restrictions, and SynthID watermarking, reflect the standard anxiety around deepfake misuse. Whether that framework proves sufficient will likely depend on how widely the feature is eventually rolled out. For now, the harder question is commercial: whether business users will treat AI self-cloning as a genuinely useful productivity tool or as a novelty that loses its appeal after the first all-hands video.
The next signal to watch is how quickly Google extends the personal avatar feature beyond its current regional restrictions, and whether the AI Expanded Access tier sees uptake as organisations weigh the cost against the creative headroom it unlocks.
