Our Story

Who We Are

Google Flower AI is Google’s dedicated creative workspace built specifically for filmmakers, visual storytellers, and independent creators. Operating within Google Labs, we are a tight-knit collective of traditional cinematographers, VFX artists, and AI research engineers. We share a core belief: a great story shouldn't be bottlenecked by a seven-figure budget, an elite production house, or years of complex rendering training.

While our roots run through Mountain View, California, our team collaborates globally—from creative directors in New York to trust and safety specialists in London. Most of us came from the world of traditional production, bringing firsthand experience with the exhaustion of long rendering queues, rigid editing software, and the frustration of having to compromise on a creative vision due to technical constraints. We built Flower because we wanted a tool that kept up with the speed of human imagination.

Since opening our platform to the public, the response has been incredible. Creators are using Flower to map out entire short films, build hyper-realistic storyboards, and bring deeply personal indie projects to life. Hearing from a director who managed to cut a production timeline in half using our tools is exactly why we show up to work every day.

How It Started

The spark for Flower began in late 2022 as an internal research experiment deep within Google DeepMind. We weren’t just trying to teach AI to recognize static pixels; we wanted it to understand the delicate physics of time, light, and emotion. We wanted it to understand how golden hour sunlight filters through a dusty window pane, or how a camera operator's subtle hand-tremble adds tension to a dramatic close-up.

The initial tests were brutally slow. Our very first internal prototype took nearly half a day to render just twenty seconds of blurry, surreal footage. But even in those early, flickering frames, we saw something undeniable: the model wasn't just generating raw shapes—it was capturing the mood we had asked for in plain English.

Over the next eighteen months, we moved the project out of pure research and built an interface around it. We brought in film editors and UX designers to transform a raw machine learning model into an actual, tactile workspace. By the time we entered closed beta, those grueling eleven-hour rendering times were down to mere seconds.

What We Build

We view Google Flower AI as a complete, living production studio—not a gimmick. We have zero interest in building a simple "text-to-video" trick that spits out random, disconnected clips. Instead, we are building an end-to-end sandbox where a creator can flesh out a rough concept, establish visual consistency across multi-scene arcs, tweak lenses and lighting setups, and assemble a cohesive rough cut.

Powering the engine under the hood is Veo, Google DeepMind’s most advanced generative video model. Because Veo was trained to interpret cinematic grammar, it responds beautifully to technical film terms like "dolly zoom," "chiaroscuro lighting," or "shallow depth of field." Wrapped around this model are our custom-built tools for real-time collaboration, allowing writing partners and editors to tweak scenes simultaneously from opposite sides of the world.

Our Commitment to Responsibility

Generative video is incredibly powerful, which means treating ethics as an afterthought is simply not an option. For us, safety isn't a compliance box we check off before a release—it is a fundamental design principle. We are fully committed to protecting the integrity of real people and safeguarding the livelihood of human artists.

Every single frame generated within Flower is instantly embedded with a SynthID digital watermark at the foundational model level. This identifier remains completely intact and detectable even if the video file is later compressed, cropped, re-encoded, or color-graded. We believe transparency in synthetic media is non-negotiable.

Our platform maintains a strict, zero-tolerance policy against the creation of non-consensual content, deepfakes designed to deceive or defame, and deliberate political misinformation. We back this policy up with multi-layered, automated safety filters and real-time human review queues. Furthermore, we are actively participating in industry-wide roundtables regarding ethical training data, fair attribution, and the long-term rights of the creative community.

The Community Driving Us Forward

The core Flower team consists of roughly 200 engineers, designers, and policy advocates. However, the true soul of the platform belongs to our global community of creators. Through our extended beta programs, creator Discord channels, and regular focus groups, we rely on independent animators, documentarians, and visual artists to tear our tools apart and tell us what actually works on a real set.

We also run the Google Flower Creator Fund, which provides direct financial grants to independent filmmakers working on non-commercial passion projects. Seeing films built with the help of Flower screen at major international festivals like Sundance and SXSW proves to us that AI can act as a bridge to independent distribution, rather than a barrier.

What Comes Next

We are just getting started. Right now, our engineering teams are tackling native, context-aware audio generation—allowing the workspace to generate ambient room tone, foley, and temporary scores that perfectly match the visual rhythm of your edit. We're also rolling out hyper-precise character consistency tools, so a character’s face, clothing, and proportions hold perfectly stable across a complex, multi-shot narrative.

Ultimately, our goal is to completely blur the line between traditional production and digital generation. Whether you start with a video shot on your phone, a mood board of still photos, or a blank script page, Flower is evolving to ensure that the technical workflow never gets in the way of the story you are trying to tell.

If you're ready to reshape the future of film with us, you're in the right place.

Get in Touch

Whether you are a creator with feedback, a researcher wanting to collaborate, or a journalist looking for a demo, we would love to hear from you.

General Enquiries: [email protected]

Press & Media: [email protected]

Creator Fund & Grants: [email protected]

Address: Google Flower AI, 1600 Amphitheatre Parkway, Mountain View, CA 94043, United States