Google's Gemini Video Model Extends to 15 Seconds with 30% Price Cut

Google DeepMind has publicly released the video generation model 「Gemini 1.2 Flash」. The model, which was previously available only to a limited set of developers, has now transitioned to a stage where it can be widely utilized. The model has undergone significant changes in both the maximum video generation time and pricing, positioning this as a move toward broader adoption of AI video generation tools.

In the field of AI video generation, the length of clips that can be generated is a critical metric that determines practical utility. Overly short videos have limited use as editing material. With this update extending the maximum generation time to 15 seconds, it is now possible to handle lengths closer to actual creative production work, such as advertisements and social media content.

On the pricing front, the generation cost at 720p (a standard high-definition resolution) has been set at $0.07 per second. This represents a 30% price reduction compared to the previous model. Additionally, by using the approach of generating a draft at lower resolution and converting only the final video to high resolution, costs can be reduced even further.

There are also changes in usability. The model now supports conversational step-by-step editing, allowing users to make gradual adjustments to video by providing text-based instructions. Rather than generating once and finishing, this approach—where fine adjustments can be made through dialogue—has the potential to make video production workflows more intuitive.

AI video generation is a field that is developing rapidly following still image generation. This release, featuring three simultaneous improvements—extended generation time, price reduction, and enhanced usability—can be viewed as a move aimed at adoption by a broader range of creators, not just a subset of advanced users. Changes to the cost structure will lower the barriers for startups and individual creators to incorporate AI video into their workflows.

The key point to watch going forward is how this price competition will ripple through other AI video generation services. Video generation AI is currently a competitive field with multiple companies, and a price cut by one company tends to influence the overall market price level. The public release of Gemini 1.2 Flash can be seen as establishing a new benchmark in that competition.