In a notable development during the Google I/O 2025 conference, Google unveiled a new AI tool for film production named "Flow."
This tool is specifically aimed at content creators and artists within the video domain.
The technology utilizes AI to streamline the process of producing short visual clips, while simultaneously offering capabilities that emulate cinematic editing tools.
Three Advanced Technologies Powering Flow
Flow's foundation rests on three technologies developed internally by Google:
Veo for video generation.
Imagen for image creation.
Gemini for understanding text instructions and guiding the model.
Collectively, these models work together to provide a comprehensive visual production experience.
Users can either input pre-existing characters or scenes, or generate them from scratch directly within the tool.
Advanced Tools for Broader Creative Control
Flow does more than just produce clips; it also provides a suite of tools for precise scene control. Among its most prominent features are:
The ability to change camera angles within a scene.
An editor for assembling and extending shots.
A tool for organizing assets and coordinating elements.
During a live demonstration of the tool, an animated scene unfolded where the camera began by zooming into a television screen, then moved inside the room, and subsequently exited through the window to show a car crossing the road.
The scene appeared seamless in its transitions, which highlights the tool's capabilities in executing complex movements.
A Platform for Sharing Works and User Ideas
Concurrent with the tool's launch, Google introduced a platform called Flow TV. This serves as a space for showcasing video clips created by others using Flow.
Notably, each clip appears alongside the prompt text used to generate it, which aids in the understanding of various approaches adopted by users.
Despite the presence of similar tools in the market, such as Moonvalley, D-ID, and Hedra, this initiative suggests that Google is no longer content with merely providing models.
Instead, it seeks direct entry into the application interface, actively competing with startups in this field.
Flow Availability and Subscription Plans
Flow is now available within the United States, as part of Google AI Pro and Google AI Ultra packages.
Subscribers to the Pro plan receive 100 video generations per month.
Meanwhile, Ultra subscribers gain access to a larger, as-yet-unannounced quantity, along with early access to the Veo 3 model, which produces video with integrated audio, including dialogue and effects.
Significant Development in Companion Models
Alongside Flow, Google announced important updates to its other models:
Veo 2 adds features such as object removal and more precise camera control.
Veo 3 was revealed, generating higher quality video and audio than Veo-2, with a deeper understanding of longer instructions.
- Imagen 4 was launched, becoming the company's best model for generating images that include clear, exportable text in various formats.
