At Google I/O 2025, the tech giant officially unveiled Veo 3, its most advanced video generation model to date, pushing the boundaries of generative AI by introducing native audio integration for the first time.
Key Features of Veo 3:
-
Generates high-quality videos from text or image prompts
-
Includes synchronized audio, such as spoken dialogue or ambient sound, generated in tandem with the visuals
-
Accurate lip-syncing and nuanced understanding of real-world physics
-
Supports storytelling, character animation, and dialogue-based scenes
-
Google claims Veo 3 can turn “a short story in your prompt” into a realistic, engaging video
Viral Examples:
Following the announcement, social media lit up with impressive demonstrations:
-
A stand-up comedy video generated entirely by AI, including a joke delivered by a digital comedian, went viral. It was created from a prompt as simple as:“A man doing stand-up comedy in a small venue tells a joke (include the joke in the dialogue)”
-
Another standout was a video of Pythagoras explaining his theorem, complete with era-appropriate visuals and synchronized voice narration. Both the video and audio were natively created by Veo 3, without separate dubbing or voice generation tools.
These videos have fooled many viewers into believing they were real, showcasing the realism Veo 3 can now achieve—especially in syncing facial expressions, lip movements, and ambient soundscapes with AI-generated speech.
Availability:
Currently, Veo 3 is in limited rollout, with access restricted to:
-
Gemini AI Ultra subscribers in the United States via the Gemini app and Google Flow
-
Enterprise users via Google Cloud’s Vertex AI platform
The Gemini Ultra subscription costs $249.99/month (around ₹21,000), which places it squarely in the professional or enterprise category rather than casual consumer use.
Unfortunately, users in India and most other regions do not yet have access to Veo 3, although wider availability is expected over the coming months as Google expands the rollout and gathers feedback.
Outlook:
Veo 3 positions Google at the forefront of AI video generation, potentially competing with models like OpenAI’s Sora. Its ability to natively generate both visuals and audio with contextual storytelling sets a new standard in multimedia creation. While access is currently limited, the performance seen in early demos suggests transformative potential in areas like education, entertainment, and marketing.
