News in Short

• Google has introduced a new AI model focused on video generation and conversational editing.
• The model can create videos using text, images, audio, and existing video references as inputs.
• It is rolling out to Gemini app subscribers, Google Flow users, and YouTube Shorts creators.
• Google says the AI model understands physics, storytelling, and contextual reasoning to produce more realistic videos.

Google has officially launched Gemini Omni Flash, its newest multimodal AI video generation model designed to create and edit videos using natural conversation. The launch marks a major expansion of the Gemini ecosystem as Google pushes deeper into AI-powered creativity.

The new AI model combines text, images, video, and audio inputs into one workflow. Users can generate cinematic clips, transform existing footage, and refine scenes across multiple prompts without restarting projects. Google says the model also maintains character consistency, scene continuity, and realistic motion during edits.

What Is Gemini Omni Flash?

Gemini Omni Flash is the first model in Google’s new Omni family. According to Google DeepMind CTO Koray Kavukcuoglu, the system combines Gemini’s reasoning abilities with advanced creative generation tools.

Unlike traditional AI video generators that mostly depend on text prompts, Gemini Omni Flash supports multiple forms of input simultaneously. Users can upload images, provide voice references, add video clips, and describe scenes in natural language. The AI then generates a connected video output.

Google says the goal is to make video creation feel more conversational instead of technical. That means users can edit scenes through follow-up prompts instead of manually adjusting timelines and layers.

For example, users can ask the AI to turn sculptures into bubbles, transform mirrors into liquid surfaces, or completely change the atmosphere of a scene while preserving continuity.

How Does Gemini Omni Change AI Video Editing?

One of the biggest talking points around Gemini Omni is its conversational editing system.

Instead of restarting prompts repeatedly, users can refine projects across multiple interactions. Google says every instruction builds on previous edits, helping maintain visual consistency and narrative flow.

This matters because continuity has been a major weakness in AI-generated videos. Characters often change appearance between scenes. Motion can also break physics rules. However, Google claims Gemini Omni improves realism by understanding gravity, kinetic energy, and fluid behavior more accurately.

The company showcased examples including chain-reaction marble tracks, claymation explainers, retro-futuristic walking sequences, and animated skateboard effects generated from simple prompts.

Google also says Gemini Omni can create educational explainers by visually breaking down complex subjects such as protein folding. This positions the model beyond entertainment and into education, social content, and communication workflows.

Why Is Gemini Omni Important in the AI Race?

The launch arrives as AI video competition intensifies across the tech industry.

Companies including OpenAI, Runway, Adobe, and Meta are rapidly building multimodal AI tools that can generate realistic video content. However, Google is trying to differentiate Gemini Omni through contextual reasoning and integrated ecosystem support.

Instead of functioning as a standalone generator, Gemini Omni connects directly with products like the Gemini app, Google Flow, YouTube Shorts, and YouTube Create.

That integration could significantly expand adoption because creators already use these platforms daily. Meanwhile, YouTube Shorts users will reportedly gain access to Gemini Omni features at no cost starting this week.

The launch also reflects a broader shift in generative AI. Companies are now moving beyond static image creation toward fully interactive media generation systems capable of understanding context, continuity, and storytelling.

Can This AI Model Create Videos From Any Input?

Google says Gemini Omni was built to support “anything from any input,” starting with video generation.

Users can combine reference images, existing video clips, music, voice prompts, and written instructions into a single generation workflow. The AI then merges those inputs into cohesive scenes.

At launch, audio input support focuses mainly on voice references. However, Google says broader audio capabilities will expand later.

The company also introduced avatar-based generation features. Users can create digital versions of themselves using their own voice and appearance for AI-generated video creation.

Still, Google acknowledged that advanced speech editing and voice manipulation features remain under testing due to safety concerns.

What About AI Safety and Watermarking?

As realistic AI video tools become more powerful, misinformation concerns are growing.

Google says every video generated using Gemini Omni includes SynthID watermarking technology. The watermark is designed to remain imperceptible while still allowing verification through Google Search, Gemini, and Chrome tools.

The company says the goal is to improve transparency around AI-generated content and help users identify edited or synthetic media online. That could become increasingly important as AI-generated clips spread across social platforms and short-form video apps.

Who Can Use Gemini Omni Flash?

Gemini Omni Flash is rolling out globally to Google AI Plus, Pro, and Ultra subscribers through the Gemini app and Google Flow. Google also confirmed that YouTube Shorts and YouTube Create users will begin receiving access this week at no additional cost.

The rollout suggests Google wants Gemini Omni to become both a creator tool and a mass-market AI product rather than a limited experimental platform. As AI video creation moves into mainstream apps, Gemini Omni could become one of the most closely watched launches in Google’s AI strategy this year.

For now, Gemini Omni Flash signals one thing clearly: AI video generation is moving beyond prompts and becoming a fully conversational creative system.