There's no way to get away with it at this point, but AI is everywhere and I've embraced more. One of the coolest new AI features that I tripped over my Galaxy Z Flip 7 has to do with the Google Gemini app and the way it generates short videos for me.
During Google I/O 2025, the company introduced an updated AI video generation model called VEO 3. What's really worth noting about this AI tool is that it can generate high quality videos simply by providing an explanation of what Gemini needs.
As a videographer and video editor, I can definitely see this in so many ways by helping me create B-roll shots that can be used. There are also other neat things that Veo 3 can do, such as converting to video using existing photos. This is what I was able to generate.
VEO 3: Things you need to know
Before you share what you can generate with Veo 3, know that you have a Google One subscription that bundles with your Gemini Pro. This allows you to access VEO 3 via the Gemini app.
Like I said, you can create a video by providing an explanation to the Gemini app, or you can upload an existing still photo and convert it to a video. It takes about a minute or two for Gemini to create a clip. The clip is 720p at 24 frames per second. It may not be high resolution compared to today's ubiquitous 4K, but these are great for sharing on social media.
All videos produced by Veo 3 Fast include both background music and sound effects, if appropriate for the clip. It was a lot of fun asking Gemini to make these clips. I realized I reached my daily limits. Hopefully this will change, but you will see that this limit is different about what kind of subscription you have.
Running on the beach
For my first test, I asked Veo 3 to take an existing photo and told Gemini to “match it like a person runs on the beach.” Although the details are a little too soft for my taste, I'll admit that it looks pretty legal, but at least the video looks accurate.
From my backyard, I really like how the camera pans and make the beach look right there. Even if the camera pans to my profile, I like the way that not only properly generates sunglasses and hats, but also adds the sound of waves hitting the sand and my running stride.
Astrophotography telescope
I don't know what happened in this second clip, but I think they should have started a new chat as they appear to think they are referring to previous descriptions of running on the beach. Using a photo I took of my Unistellar Odyssey Pro Smart Telescope in my backyard, it instantly jumps to another completely different shot of a person running on the beach at night.
I asked Gemini to “make it so that there are auroras in the sky.” Despite the confusion that is partly my fault, it produces the video. Just like before, the details are a little softer than what I would like, but at least the animation looks right.
I tried it again in a fresh chat with Gemini, but this time it produced exactly what I wanted. What's interesting about this video generation is how to apply ambient light to the rest of the scene and give it a little animation to the surrounding trees and branches.
Alien invasion
This really pops out, mainly about how I lay out the scenes from the photos I provided. In a simple shot of the park I took with my cell phone, I asked Gemini to “make it so that alien invasions happen on the scene.”
It starts with a shot from my photo, but the camera quickly moves to the park's playground with a saucer flying overhead with a laser beam. Without looking at the details, you might miss some of the weirdness of the video. Nevertheless, it is an example of how a simple (and quick) VEO 3 can easily (and quickly) create videos from a simple description.
Cat moving to music
I took a photo of the black cat on Sony headphones and asked Gemini to “put the cat on the table and rock the music.” The only movement applied to the photo is on my cat, but I can see how the headphones wobble a little while as he moves around.
Unfortunately, I failed to get my headphones to wear. Apart from that, it is the most detailed and compelling video in my opinion.
Chester Bennington portrait
Up to this point, I have an existing gemini photo. I've captured it to create a video that speeds up veo3. However, on this next test, I downloaded a portrait of former Linkin Park lead singer Chester Bennington and asked him to “put him in a boxing ring fighting someone.”
I'm even surprised that it allowed me to take existing photos of celebrities, but it can't do the rest. The only thing that ultimately does is apply a bit of animation to Chester's photos and fail to put him in the boxing ring. However, the audio has a ringside announcer speaking via video to showcase the boxer.
Optimus Prime vs Donatello
For the final request, I asked Gemini to generate a video from scratch that was nothing more than a text description. So I didn't upload any photos at my request. At that time I asked him to “make a video from Teenage Mutant Ninja Turtles about Transformus Prime fighting Donatello.”
This looks more like an animated video than anything else, but it created exactly what I wanted. Not only are they squared, but Gemini are intelligent enough to know who Donatello is with the other ninja turtles. I also like the special effect of the Optimus Prime blaster hitting Donatello.
VEO 3: Bottom Line
I've never tried any AI video generator tools, but this is definitely impressive. There's still work to remove some of those weirdness in some clips, but I have to say it's a quick and easy way to create custom videos.
In addition to the visuals, I think Veo 3 will accurately apply sound effects to make it feel more reliable. I'm curious to see what Google is doing in the end. This includes whether or not there are other subscriptions that offer users three or more videos per day.
to follow Tom's Guide on Google News To get the latest news, how-tos and feed reviews. surely[フォロー]Please click the button.
