Google is developing a new “Video Remix” feature for Google Photos, internally codenamed “Soba,” which aims to bring generative AI-powered editing to personal video clips. According to findings by app forensic specialist AssembleDebug, the feature will likely allow users to transform existing videos using AI, mirroring the functionality of the current Photo Remix tool. While the feature remains in early development, code indicates it will use a dedicated interface, potentially leveraging Google’s Gemini Omni model to apply creative effects to video footage.
How will the Soba feature change Google Photos?
Evidence discovered in the Google Photos Android app shows that “Soba” functions as a direct successor to the existing image-editing tools. According to AssembleDebug, activating the Soba code triggers a rename of the current “Remix” button to “Photo Remix.” This naming convention suggests that Google intends to distinguish between still-image edits and the upcoming video-based transformations. The new tool features a YouTube-style play icon combined with Google’s signature “sparkle” Gemini logo, signaling that the feature is designed to process moving footage rather than static photos.
What technology powers Video Remix?
The transition from photo to video editing likely involves shifting from Google’s Nano Banana model to the more capable Gemini Omni. According to reports, Gemini Omni is Google’s multimodal model designed for conversational editing and context-aware transformations. Unlike simple filters, this technology can theoretically interpret natural language prompts to change the style of a video—such as applying a claymation aesthetic—while maintaining the integrity of specific events within the clip. However, current data suggests that Google may restrict these capabilities to a set of pre-defined presets, similar to how the initial Photo Remix launch limited users to 13 specific styles.
How does this compare to existing Google Photos tools?
Google currently offers a “Photo to video” feature, but the Soba project represents a fundamental shift in processing. While existing tools create new clips from still images, Soba is designed to ingest and modify existing video files. The following table highlights the expected differences between the current and upcoming workflows:
| Feature | Input | Primary Model |
|---|---|---|
| Photo Remix | Still Photos | Nano Banana |
| Soba (Video Remix) | Video Clips | Gemini Omni (Expected) |
Frequently Asked Questions
Will Video Remix be free for all users?
Google has not announced pricing for Soba. However, previous AI features in Google Photos have often been tied to Google One subscriptions or specific hardware tiers.
Can I use Video Remix to edit my existing videos right now?
No. According to AssembleDebug, the necessary code to execute the Soba feature is not yet present in the Google Photos application. The current files are only preparatory assets.
Does Video Remix use generative AI?
Yes. The use of the Gemini sparkle icon strongly indicates that the feature relies on generative models to create new visual content based on your original video input.
Are you excited to see AI take over your video editing tasks, or do you prefer manual controls? Let us know your thoughts in the comments below, and subscribe to our newsletter for the latest updates on Google’s AI roadmap.
