Google Gemini May Soon Add Video Generation Capabilities, APK Teardown Reveals

Gemini could soon be upgraded with video generation features (APK teardown)

We couldn't spot more code, but we can join the dots and speculate on the upcoming functionality. "Videogen" most likely refers to video generation, and this code most likely points at video generation features within Gemini. Google offers AI-powered video generation features through Google Vids, but it's not a generative AI system. Instead, Google Vids helps build a video from a simple idea by walking you through each step of the process, including ideation, script writing, layout design, editing, and more. You can create a first draft video, remove image backgrounds, generate AI images, generate a voiceover, and record a studio read-along teleprompter, all within Google Vids. Such a system could feel at home within Gemini.

Google might be exploring AI video generation for Gemini

Gemini Live could soon recap conversations automatically before you exit Summary Code discovered in the Google app hints at upcoming video generation capabilities for Gemini, Google's AI assistant. This could significantly expand its functionality beyond image creation. It's unknown whether Gemini's video generation will resemble Veo 2 (fully generative, like Sora) or Google Vids (AI-assisted production). The code provides no definitive answer. While "videogen" is mentioned in the app's code, the feature is likely still in development. The release timeline and specific model used remain unclear. Google Gemini can help you build videos using prompts, but this functionality is exclusively limited to the tech giant's Vids productivity tool. However, the exclusivity might soon end. Soon after bringing Veo 2's text-to-video generation capabilities to YouTube Shorts, it looks like the Gemini AI assistant might be next in line to help you with your next video project. Related Google DeepMind's version of Sora and Adobe Firefly is coming to YouTube Shorts Make short work of Shorts Posts For reference, the AI tool currently uses Imagen 3 to generate images from prompts, and according to Android Authority, it might soon use a dedicated model to generate videos too. What model the feature might use is currently unclear, and so is its release timeline. However, despite the unknowns, the explicit mention of "Videogen" within version 16.6.23 of the Google app makes it clear that Google is exploring and/or actively building the functionality. References to Videogen appear alongside "Robin," which is the given codename for Gemini features within the Google app, alongside strings like ""Working ......" and "We will notify you when it's ready," which are likely the prompts that the Gemini app will show you when it is rendering your video. This could make Gemini a true all-in-one AI tool Source: Android Authority The feature's integration doesn't appear to be anywhere near done, and it is unclear if the video generation capabilities it will offer will be more in-line with Veo 2 or Google Vids. For reference, the former is similar to OpenAI's Sora, wherein the tool generates a completely animated video with realistic motion and complex scenes from scratch. Vids, on the other hand, offers AI-assisted video production in the form of slides, based on the data and rough ideas you share with it. We'll likely learn more about which one of the two implementations Gemini might lean towards if and when it is close to launch. Related Google Vids is rolling out now, making video creation a breeze for Workspace users You might need to update your browser though Posts