Google updates its AI video editor Vids with Gemini Omni and adds a feature that lets users create a custom digital avatar using a selfie and a voice recording
OpenAI's Sora may have shut down, but Google apparently thinks there's still interest in a tool that lets you star in your own AI videos.
Context & Ripple Effects
Vids has been moving from a basic video-editing offering toward AI-assisted production: earlier coverage added stock AI avatars, transcript trimming, image-to-video generation, prompt-directed avatar controls, Veo support, and YouTube export.
Gemini Omni was introduced as Google’s multimodal creation model with video generation as an initial use case. Its integration into Vids brings that model layer into a product already oriented around making and distributing finished videos.
First-order effects
- Vids users gain Gemini Omni-backed creation capabilities alongside a custom-avatar workflow that uses a selfie and voice recording, lowering the effort required to put a recognizable presenter into an AI-made video.
- Google deepens the connection between its flagship multimodal model and a dedicated video-production surface, rather than leaving video generation as a standalone model capability.
Second-order effects
- AI video tools competing for business and creator workflows face greater pressure to combine generation, editing, presenter creation, and export in one product flow rather than offer isolated generation features.
- Because the new workflow uses a person’s image and voice, product differentiation will increasingly depend not only on generation quality but also on clear controls around how digital likenesses are created and used.
Third-order effects
- If this product pattern continues, AI video creation is likely to shift from specialist generation tools toward workflow-native suites that package model access, editing, avatars, and publishing together.
- The growing use of personalized avatars could make likeness, consent, and provenance controls a more consequential part of AI-content product design, particularly as such tools become easier to use.
The trend: This is one data point in the commercialization of multimodal AI through integrated content-production workflows, where generation models are embedded in end-user creative applications.