Google has introduced Gemini Omni 1.2 Flash, an updated AI video generation model designed to give creators and developers more control over how they produce, edit and refine video content.
The latest release extends video generation to 15 seconds while introducing improvements in visual consistency, conversational editing, audio quality and camera movement. Google has also lowered generation costs, making the technology more accessible to businesses producing video content at scale.
The update reflects a growing shift in generative AI. Creating a video from a text prompt is no longer the only priority. Developers and creative teams increasingly need tools that can revise existing footage, preserve product details and deliver usable results without starting from scratch.
Gemini Omni 1.2 Flash Extends AI Video Generation to 15 Seconds
Google has expanded the maximum video duration from 10 seconds to 15 seconds, giving users additional room to develop scenes and communicate ideas within a single generated clip. The model supports flexible durations between three and 15 seconds, allowing creators to adjust video length according to their production needs.
The additional runtime could be particularly useful for social media advertising, product demonstrations and short promotional content. Instead of generating several separate sequences and combining them manually, creators can produce longer scenes while maintaining continuity between subjects, objects and backgrounds.
Google has also worked on improving how the model handles movement across extended sequences. Maintaining consistent visuals becomes more challenging as videos get longer, especially when the camera changes angles or multiple objects interact within a scene.
Google Reduces AI Video Generation Costs by 30%
Cost is another major focus of the Gemini Omni 1.2 Flash update. Google has reduced video generation pricing by 30%, with 720p output priced at $0.07 per second. Based on that rate, a complete 15-second video would cost approximately $1.05 before any additional processing or related charges.
Lower pricing could make AI video production more practical for businesses that regularly create advertisements, product demonstrations and social media campaigns. Marketing teams often need several versions of the same content, particularly when testing different audiences, formats or creative approaches.
The model also supports workflows that begin with lower-resolution previews before moving to higher-quality output. Creators can experiment with different concepts, refine selected clips and reserve more expensive rendering for the final versions. This gives production teams greater flexibility when balancing creative quality against operating costs.
Conversational Editing Gives Creators More Control
One of the more practical improvements in Gemini Omni 1.2 Flash is its expanded support for conversational video editing. Users can provide natural-language instructions to modify existing footage rather than generating an entirely new video whenever they want to change a particular element.
A creator could ask the model to replace a background, adjust lighting, modify a product’s appearance or change the visual style while preserving the original composition. This approach reduces the need to recreate scenes from the beginning and gives users more opportunities to refine their work through successive instructions.
For advertising agencies and content creators, the feature could simplify the production of multiple campaign variations. A single video concept might be adapted for different markets, products or platforms without rebuilding every element manually. The usefulness of this capability will depend on how reliably the model preserves details that users want to keep unchanged.
Improved Visual Consistency Supports Product and Brand Content
Google has also focused on a persistent weakness in AI-generated video: keeping subjects and objects visually consistent throughout a scene. Gemini Omni 1.2 Flash introduces improvements intended to preserve details such as clothing patterns, product packaging, colors and character appearances as the video progresses.
This is particularly relevant for e-commerce businesses and brands using AI-generated footage to showcase physical products. An advertisement becomes difficult to use when a product changes shape, its packaging looks different between frames or its colors no longer match the original reference.
By improving reference-image handling and visual continuity, Google is attempting to make generated videos more suitable for commercial applications where accuracy matters as much as visual appeal. Fashion retailers, product marketers and creative agencies could benefit from these capabilities, although generated footage still requires careful review before publication.
More Natural Dialogue and Cinematic Camera Movement
Audio generation and camera movement are receiving improvements in Gemini Omni 1.2 Flash. Google says the model produces more natural dialogue, stronger synchronization between speech and lip movements, and better handling of scenes involving multiple speakers.
These improvements could support short educational videos, animated conversations, promotional explainers and other content that combines spoken dialogue with generated visuals. More convincing audio synchronization may also reduce the need for separate post-production adjustments.
Camera movement has become another important area of development. The model is designed to produce smoother cinematic sequences, including zooms, orbiting shots and continuous movement through environments. Rather than relying on disconnected scenes, creators can attempt more deliberate camera directions that give generated videos a stronger sense of visual storytelling.
Gemini Omni 1.2 Flash Targets Commercial Video Production
Google is positioning Gemini Omni 1.2 Flash for more than individual experimentation. Its lower generation costs, longer video duration and improved editing capabilities are intended to support commercial production workflows where businesses need to create content repeatedly and efficiently.
Creative agencies, advertising teams and e-commerce platforms increasingly use generative AI to test visual concepts, produce promotional material and develop personalized marketing assets. The ability to revise existing footage while retaining important details could help reduce the time required to prepare different versions of a campaign.
For smaller businesses, the appeal is somewhat different. AI-generated video could provide an alternative to expensive production setups for certain types of content, particularly short promotional clips and early-stage creative concepts. Traditional filming and professional editing will remain important, but generative tools are becoming another option in the production process.
Google Expands Access Through Its AI Developer Ecosystem
Gemini Omni 1.2 Flash is designed to fit into Google’s broader AI development environment, giving developers and businesses ways to incorporate video generation into their own applications and workflows.
Access through Google’s developer tools and cloud infrastructure could encourage more businesses to experiment with AI-powered video services. Instead of relying exclusively on standalone creative applications, developers may be able to build video generation and editing features directly into existing platforms.
That approach could lead to wider use of generative video in marketing software, e-commerce systems and content management platforms. The commercial opportunity is not limited to people creating videos manually. Businesses may eventually use these models to generate large volumes of customized content through automated production pipelines.
AI Video Generation Still Faces Technical Challenges
Despite the improvements, AI-generated video continues to struggle with certain types of complex movement and physical interactions. Fast action, rapidly changing scenes and detailed object behavior can expose inconsistencies that are less noticeable in slower or simpler footage.
Longer videos also introduce additional challenges. Maintaining accurate product details, realistic motion and consistent character appearances across an extended sequence requires more than generating visually impressive individual frames.
For businesses using the technology in advertising or customer-facing materials, human review remains essential. Product representations, dialogue, branding and other important details need verification before generated videos are published.
Gemini Omni 1.2 Flash represents progress in addressing these problems, but it does not eliminate the limitations of generative video technology.
What Gemini Omni 1.2 Flash Means for the Future of AI Video
Google’s latest model highlights how competition in AI video generation is changing. Longer clips and realistic visuals remain important, but affordability, editing flexibility and dependable output are becoming major factors in whether businesses adopt these tools.
A model capable of producing attractive footage is useful for experimentation. A model that allows creators to revise that footage, preserve recognizable details and generate multiple versions at a reasonable cost has a stronger chance of becoming part of everyday production.
Gemini Omni 1.2 Flash moves Google’s video generation technology further in that direction. Its combination of longer clips, lower pricing and conversational editing could appeal to developers, agencies and businesses looking for more practical ways to produce visual content.
The real test will come when creators begin using these capabilities across demanding commercial projects. Consistency, reliability and creative control will ultimately determine whether AI-generated video becomes a dependable production tool rather than simply another impressive demonstration.
Sources
1. Google Blog — Gemini Omni 1.2 Flash: Fast and Cost-Effective With Longer Clips (November 5, 2026). Official announcement covering longer videos, pricing, editing and model improvements.
Read Google’s official announcement
2. Google Cloud Blog — Gemini Omni 1.2 Flash Now Generally Available (November 5, 2026). Details on enterprise deployment, production workflows and commercial applications.
Read the Google Cloud announcement
3. Google AI for Developers — Video Generation in the Gemini API. Technical documentation covering Gemini Omni Flash and video generation capabilities.
