On November 5, 2026, Google moved Google Vids, its video editing tool, onto a new video model called Omni 1.2 Flash. With the switch it set a clear free allowance: anyone with an ordinary Google account, or a Google Workspace subscription, can generate up to three videos a month, each up to 15 seconds long, at no cost.
The tool opens in the browser at vids.new. Google's earlier announcement of the feature, in September, presented video generation as a desktop task. According to Google's help page, the AI features in Vids are available in every country where Google Workspace is sold, and Israel is one of them.
45 seconds of video a month
Three 15-second clips add up to 45 seconds a month. That will not make a brand film. It is, however, the exact length of an Instagram story, a WhatsApp status or a short Facebook video ad. A business that runs one promotion a week can cover three of the month's four weeks this way, with no photographer and no editor.
Google names this use directly: a product promo generated in a single take, without a cut, ready for social media. The announcement also says the new model makes the narration in clips sound more natural.
Starting from a real product photo
For a small business the most useful part is image-to-video. You upload reference images, and Google says characters and scenes stay consistent from the first frame to the last. In practice that means starting from an actual photo of the product on the shelf, rather than a written description the model fills in with a look of its own.
A florist can upload a photo of a bouquet and ask for the camera to circle it slowly. A Pilates studio can upload a picture of the empty room and get a short clip of light coming through the window. A maker of handmade soap can upload one bar and get a clip of it turning on a wooden board. In each case it pays to ask for one simple movement. The more complicated the request, the more likely something in the scene changes along the way.
Prompt in English, captions in Hebrew
Google's help page says many AI features in Vids are currently available in English only. The safe route is to write the prompt in English, even when the video is meant for Israeli customers. Hebrew text, such as the name of the sale or a price, is better added afterwards in the regular editor than drawn into the video by the model. Text that a model renders inside an image tends to warp, and Hebrew raises that risk.
Hebrew narration is not guaranteed either. In September Google announced narration in more than 100 languages "coming soon", and the November announcement does not say whether Hebrew is now included. Anyone who needs a Hebrew voiceover should test it before building a campaign around it.
The free allowance and the plans above it
The three videos renew every month. Users who need more are pointed to Google's paid AI plans for personal accounts, or to the business Workspace plans. According to the help page, paid plans allow up to 500 seconds of video clips a month, and some plans have a shared limit of 10 generations a month across AI avatars and video clips.
Per the September announcement, every clip made in the tool carries an invisible digital watermark called SynthID, which lets the video be identified as AI-generated. There is no visible label on the video itself.
Before it goes on your feed
A video generated by AI still shows the product the business sells, so it has to look like the real thing. If the model added flowers the bouquet does not have, or gave the packaging a color that is not in stock, the customer will notice when the order arrives. Check the clip frame by frame against the original photo, and only upload images the business holds the rights to.