Google Gemini’s Video Uploads Are So Close That Someone Has Already Enabled Them
Sign in to your Android Police account
Summary
- Google Gemini is testing video uploads, allowing users to ask questions about their videos.
- Users can upload multiple videos, but the combined length is currently limited to 5 minutes.
- The feature may be officially launched during Google I/O 2025.
Google Gemini is a great tool for answering questions, summarizing files, and even researching the web for you. However, one of its missing capabilities has been the ability to upload and analyze video files. Google has reportedly been working on this feature since early March, and now we’ve gotten our first look at the feature in action.
Related
Android Authority managed to get the video analysis feature working on Gemini for Android (version 16.15.38.sa.arm64 beta), giving us an early preview of how it might function once it rolls out more widely. The demo shows that users will be able to upload videos directly to the Gemini app using the “+” upload button. Once selected, the Android media picker will display photos and videos stored on the device.
This hands-on demo shows what Gemini can do with your videos
After a video is uploaded, you’ll be able to ask questions about the content, and Gemini will analyze the footage to provide answers. In the hands-on, the user was able to upload three separate videos, and Gemini handled each one effectively. One of the more impressive examples showed Gemini identifying the video’s location without any context.
Another takeaway from the hands-on video shows that Gemini will support uploading multiple videos at once. However, the combined length of all uploads — or of a single video — is currently capped at 5 minutes. Videos longer than that will trigger an upload error.
This isn’t the first time we’ve heard about the upcoming video analysis feature. Earlier leaks even revealed the supported video file types. Still, despite this feature now being seen in action, there’s no word on when it will officially launch. A good guess would be an announcement during Google I/O 2025 next month, as this seems like a major new feature.
I haven’t been able to test the feature myself, but I can already see the potential. Whether it’s asking questions about a video tutorial or analyzing something you screen recorded on Android, this could be a very helpful addition.
