Veo 3 - An Overview
Ambitious Part-Time Blogger. Software Engineer. Tech Enthusiast. Interested in AI and ML. Passionate about XR. Love to learn more about Cloud Technologies.
Veo 3 videos are taking the world by storm, and you would be surprised at how big the latest AI trend is. Let’s have an overview on what it is, and how we can distinguish AI from reality.
If you would prefer a video version of the content, kindly have a look at this instagram reel.
Your feeds might have been folded with a different kind of video with some quirks, and wondered, it looks real, but something is off, but you can’t really place it.
You just witnessed the latest AI Trend, with videos that are generated using the new VEO 3 model or using the updated VEO 2.
Google IO 2025 introduced the new model in May 2025, and it is a strong hit. VEO 3 models can generate audio, including dialogues and scene specific sounds, on top of the video generation capabilities of VEO 2
As the line between AI and reality blurs, a question that arises is “What is real?” and the answer to that lies in the video’s metadata. Synth ID, a tool that watermarks and identifies content generated through AI, by digital watermarks directly into AI-generated images, audio, text or videos, and it can be verified using the SynthID Detector Portal. The capabilities are still in waitlist access.
VEO 3 models are great at understanding; you can tell a short story in your prompt, and the model gives you back a clip that brings it to life. The other improvements include better reference images, camera controls, outpainting, and object management.
If you feel you need to give it a try, you can access them over Flow, Google AI Studio and all these options. You can generate videos upto 4k based on your subscription tier, with the free tier giving u 720p videos, of duration upto 8s.
A honest question that would pop up in everyone's mind is that, are we ready for such a change, and are we ready to handle a large scale information crisis? The honest answer is no. Over the last few years as image generation got a lot more realistic, we are wondering “Can this be real”, and as time moves on and the capabilities increase for the video models we would need stricter guidelines to distinguish the myth and the made.
