Google has announced two updates for its Vids video creation tool: Gemini Omni integration for AI-assisted editing and personal avatars that let users star in videos without filming themselves. These features aim to make video production faster and more accessible.
SynthePulse Insight · AI deep reading
Version 1 · 1 source
Google Vids introduces Gemini Omni multimodal generation and editing, along with a personal digital avatar feature, making video creation fully AI-driven from shooting and editing to on-camera presence. However, regional restrictions, age limits, and digital watermarking also pose new transparency challenges.
On July 16, 2026, Google Vids announced two major updates, centered on introducing the Gemini Omni model. This model enables users to generate high-quality video clips using natural language text prompts and image references (such as photos or sketches). Unlike earlier versions, Omni can blend multiple input modalities, making outputs more aligned with user intent.
More critically, Omni supports step-by-step editing: after generating an initial draft, users can ask Vids in everyday language to change the background, adjust lighting, or add effects, without starting from scratch. This capability also applies to users' own footage, marking a shift from parameter-based editing to conversational interaction.
Previously, Google opened the Veo 3.1 model to all users in February 2026, driving the adoption of video generation within Vids. Official data shows that over the past year, users have created millions of videos in Vids.
The personal digital avatar feature allows users to upload a selfie and a short voice recording, and the system generates an AI avatar that resembles the user in appearance and voice. Users simply input the text they want to express, and the avatar automatically 'performs' and speaks the content, eliminating the need for actual filming or recording.
This feature is currently available only to Google AI Pro and Ultra subscribers, as well as Google Workspace business customers. Additionally, digital avatar usage is region-restricted, limited to users aged 18 and older in specific areas. Each digital avatar is tied to the user's Google account and is limited to that account holder's likeness to prevent misuse.
To address the authenticity challenges of AI-generated content, Google embeds an invisible SynthID digital watermark in every video clip generated by Gemini Omni or digital avatars. This watermark allows others to verify whether the video was AI-generated, thereby maintaining content transparency while encouraging creative expression.
This mechanism is part of Google's ongoing investment in AI content safety. Although watermarking technology aids traceability, its ease of detection and adoption depend on third-party tool support. It is currently unclear whether the watermark remains robust after compression or secondary editing.
Gemini Omni and digital avatar features are integrated into Google Vids but are not available to all users. Conditions for using personal digital avatars include: Google AI Pro/Ultra subscription or Workspace business customer, being at least 18 years old, and being located in a supported region. These restrictions may stem from regulatory compliance (e.g., minor protection) and computational resource allocation.
Notably, the strong binding of digital avatars to users' Google accounts means their usage scenarios are strictly limited, preventing cross-account or anonymous use. This somewhat restricts flexibility but also reduces the risk of identity theft.
The information in this article primarily comes from Google's official blog, constituting first-hand product announcements. All feature descriptions, limitations, and data are from official statements, but actual user experience and performance require independent verification. The robustness of SynthID watermarks and the availability of detection tools were not detailed in the announcement.
Google Vids, through Gemini Omni and digital avatars, lowers the barrier to video creation to 'just typing,' but regional, age, and subscription restrictions mean this capability is currently available only to specific user groups. SynthID watermarks provide verifiable transparency for AI content, but their actual effectiveness remains to be seen.
Primary report
Primary source
Also covered by 1 sources