Google DeepMind
Build with Gemini Try Gemini

Veo

Our state-of-the-art video generation model

Try in Gemini Try in Flow

Introducing Veo 3, our video generation model with expanded creative controls – including native audio and extended videos.

  • Re-designed for greater realism

    Greater realism and fidelity, made possible by Veo 3’s real world physics and audio.

  • Follows prompts like never before

    Improved prompt adherence, meaning more accurate responses to your instructions.

  • Improved creative control

    Offers new levels of control, consistency, and creativity – now across audio.

Introducing Veo 3.1

Video, meet audio. Our latest video generation model, designed to empower filmmakers and storytellers.

Learn how to prompt

Prompt: A medium shot opens on a seasoned, grey-bearded man in sunglasses and a paisley shirt, his gaze fixed off-camera with a contemplative expression. His gold chain glints subtly. Beside him, a younger man in a tank top, also looking forward, suggests a shared moment of observation or reflection. The camera slowly pushes in, subtly emphasizing their quiet focus. In the background, a vibrant mural splashes across a wall, hinting at an urban setting. Faint city murmurs and distant chatter drift in, accompanied by a mellow, soulful hip-hop beat that adds a contemplative yet grounded atmosphere. "The city always got a story," the older man murmurs, a slight nod of his head. "Just gotta listen."

Veo 3 lets you add sound effects, ambient noise, and even dialogue to your creations – generating all audio natively. It also delivers best in class quality, excelling in physics, realism and prompt adherence.

Prompt: A medium shot frames an old sailor, his knitted blue sailor hat casting a shadow over his eyes, a thick grey beard obscuring his chin. He holds his pipe in one hand, gesturing with it towards the churning, grey sea beyond the ship's railing. "This ocean, it's a force, a wild, untamed might. And she commands your awe, with every breaking light"

Prompt: A follow shot of a wise old owl high in the air, peeking through the clouds in a moonlit sky above a forest. The wise old owl carefully circles a clearing looking around to the forest floor. After a few moments, it dives down to a moonlit path and sits next to a badger. Audio: wings flapping, birdsong, loud and pleasant wind rustling and the sound of intermittent pleasant sounds buzzing, twigs snapping underfoot, croaking. A light orchestral score with woodwinds throughout with a cheerful, optimistic rhythm, full of innocent curiosity.

A wise old owl and a nervous badger sit on a moonlit forest path. "They left behind a...a 'ball' today. It bounced higher than I can jump.” the badger stammered, trying to comprehend it. “What manner of magic is that?" the owl hooted thoughtfully. Audio: Owl hooting, badger's nervous chitters, rustling leaves, crickets.

A wise old owl flies away out of the frame and a nervous young badger runs in a different direction out of the frame. In the background, you can see a squirrel hurrying past making noise of rustling dried autumn leaves as it goes. Audio: birdsong, loud and leaves rustling, and the sound of intermittent pleasant sounds buzzing, twigs snapping underfoot, and the sounds of squirrels scurrying through the dried autumn leaves. The sound of an owl hooting in the distance, badger's nervous chitters, rustling leaves, crickets, sounds that are full of innocent curiosity.

Prompt: A medium shot, historical adventure setting: Warm lamplight illuminates a cartographer in a cluttered study, poring over an ancient, sprawling map spread across a large table. Cartographer: "According to this old sea chart, the lost island isn't myth! We must prepare an expedition immediately!"

Prompt: A detective interrogates a nervous-looking rubber duck. "Where were you on the night of the bubble bath?!" he quacks. Audio: Detective's stern quack, nervous squeaks from rubber duck.

Prompt: A close up of spies exchanging information in a crowded train station with uniformed guards patrolling nearby "The microfilm is in your ticket" he murmured pretending to check his watch "They're watching the north exit" she warned casually adjusting her scarf "Use the service tunnel" Commuters rush past oblivious to the covert exchange happening amid announcements of arrivals and departures

Prompt: Camping (Stop Motion): Camper: "I'm one with nature now!" Bear: "Nature would prefer some personal space."

Prompt: The scene explodes with the raw, visceral, and unpredictable energy of a hardcore off-road rally, captured with a dynamic, almost found-footage or embedded sports documentary aesthetic. The camera is often shaky, seemingly mounted inside one of the vehicles or held by a daring spectator very close to the action, frequently splattered with mud or water, catching unintentional lens flares from the natural, often harsh, sunlight filtering through trees or reflecting off wet surfaces. We are immersed in a challenging, untamed natural environment – perhaps a dense, muddy forest trail, a treacherous rocky incline littered with loose scree, or a series_of shallow, fast-flowing river crossings. Several heavily modified, entirely unidentifiable, and unbranded off-road vehicles are engaged in a frenetic, no-holds-barred race. These are not showroom models; they are custom-built, rugged machines – open-wheeled buggies with exposed engines and prominent roll cages, heavily armored pickup trucks with oversized, knobby tires and snorkel exhausts, their original forms and manufacturers completely obscured by extreme modifications, layers of caked-on mud, and a general air of brutal functionality. The dominant sounds are the deafening, guttural roar of powerful, untamed engines, the whine of transmissions, the percussive impact of suspension bottoming out, and the constant spray of mud and water. Within an 8-second sequence, one of the lead vehicles, a low-slung, open-cockpit buggy so caked in thick, brown mud that its original color is a mystery, approaches a wide, shallow river crossing at incredible speed. Without the slightest hesitation, its unseen driver powers straight into the water. The impact sends an enormous, almost solid, opaque sheet of muddy water, mixed with stones and debris from the riverbed, spectacularly high into the air, completely engulfing the small buggy for a terrifying moment, obscuring it from view as if it has been swallowed by the river itself. Right on its tail, a pursuing, equally mud-encrusted, custom-built truck – a hulking, high-clearance beast with a heavily reinforced external roll cage and no discernible badging – arrives at the river crossing just as this massive wall of airborne water reaches its peak. Instead of slowing or attempting to find a clearer path, the truck's driver, with unwavering aggression, plunges directly into and through this opaque, turbulent curtain of muddy spray at full throttle. A split second later, the truck bursts out from the other side of the deluge, water cascading from its roof and chassis, its oversized windshield wipers struggling frantically to clear the torrent of muddy water obscuring the driver's vision. It lands heavily on the far bank, suspension groaning, but still in hot pursuit of the now-reappearing buggy. This thrilling, messy, and visually spectacular sequence of one vehicle creating a massive environmental obstacle and the next immediately conquering it through sheer force, forms the core, immersive, attention-grabbing event of the 8-second sequence. The race continues with undiminished ferocity, the natural terrain itself an active participant in the conflict.

Prompt: A meticulously detailed scene opens, displaying a small, pale yellow, humanoid figure crafted from wax. This figure stands centered in a warm, ethereal landscape composed entirely of molten wax, which forms gently undulating hills and reflective pools. In its raised hand, a delicate, bright flame flickers with a vibrant glow, casting soft, warm light on the figure's smooth, slightly reflective surface. To the left, a larger, partially melted candle drips viscous wax onto a nearby mound, its own blue-tinged flame barely visible. The atmosphere is serene, illuminated by the golden light of the small figure's flame, highlighting the glossy textures and subtle translucence of the wax environment. (0-1 seconds) The camera initiates a smooth, tracking shot, maintaining an eye-level perspective with the small wax person. As the figure begins to gently walk forward, its small feet creating subtle ripples in the viscous, pale yellow wax terrain, the camera gracefully follows its movement. The figure takes slow, deliberate steps across the shimmering, honey-colored landscape, its arm steadily raised to protect the precious, unwavering flame. Each step is deliberate, conveying a sense of purpose. The soft glow of the flame remains the primary light source, illuminating the path ahead and emphasizing the intricate, dripping textures of the surrounding wax formations. (1-7 seconds) The wax person continues its quiet journey, steadily progressing across the glowing, soft landscape. The camera holds its smooth, tracking motion, subtly receding slightly to reveal a broader view of the wax world, emphasizing the figure's determined, solitary walk through its unique environment. The flame continues to burn brightly, a beacon in the warm, diffused light. (7-8 seconds)

Prompt: The scene opens with a top-down or wide-angle shot showcasing a vast, perfectly flat, neutral-colored surface – perhaps the polished concrete floor of an enormous, empty aircraft hangar, or a giant, minimalist tabletop stretching beyond the frame, under bright, even, shadowless studio lighting. This surface is meticulously covered with thousands upon thousands of small, identical, brightly colored paper squares, arranged in a simple, orderly grid. Each square is a single, vibrant, uncreased sheet – a sea of reds, blues, yellows, greens, oranges, creating a stunning, static mosaic of pure potential. The atmosphere is one of quiet anticipation, a sense of immense latent energy waiting to be unleashed. There is no visible mechanism, no hint of how these papers might be manipulated. Within an 8-second sequence, initiated by an unseen cue – perhaps a subtle, almost inaudible, low-frequency hum that ripples almost invisibly across the surface, or a sudden, soft flash of diffused light – all the thousands of paper squares simultaneously, and with breathtaking precision, leap a few inches into the air as if startled into life. Then, in a mesmerizing, perfectly synchronized, and incredibly high-speed aerial ballet, they begin to fold themselves in mid-air. With impossible, almost magical celerity and accuracy, unseen forces guide each individual square through a complex series of sharp creases, neat tucks, and intricate folds. The swarm of fluttering, self-constructing papers is a blur of color and motion, a chaotic yet utterly controlled explosion of activity. Within a mere five to six seconds, this frenetic process of airborne origami completes. Each of the thousands of squares has transformed into an identical, perfectly formed, complex origami figure – perhaps graceful cranes with outstretched wings, delicate multi-petaled lotus flowers, or miniature, intricately detailed dragons. In the final two to three seconds of the sequence, these newly formed origami figures, still hovering in mid-air, then smoothly and rapidly arrange themselves, like a flock of perfectly trained birds or a sophisticated, self-organizing swarm of nanobots, into a stunning, larger, three-dimensional collective pattern or a recognizable mosaic image – perhaps a giant, hovering sphere composed of countless tiny birds, or a complex, flowing wave of flowers, or even a pixel-perfect, three-dimensional representation of a face or symbol. This collective sculpture holds its form for a beat before the individual origami figures begin to gently, gracefully, and silently settle back down onto the surface, now arranged in their magnificent new configuration. This entire rapid, impossible, and beautiful transformation – from simple squares to a synchronized swarm of self-folding forms creating a complex collective artwork – is the core, eye-popping, and meticulously detailed VFX spectacle of the 8-second sequence. The visual is one of magical precision, emergent complexity, and the beauty of mass synchronized action.

Prompt: In rural Ireland, circa 1860s, two women, their long, modest dresses of homespun fabric whipping gently in the strong coastal wind, walk with determined strides across a windswept cliff top. The ground is carpeted with hardy wildflowers in muted hues. They move steadily towards the precipitous edge, where the vast, turbulent grey-green ocean roars and crashes against the sheer rock face far below, sending plumes of white spray into the air.

Prompt: The camera slowly pushes forward into a breathtaking ice cave, its jagged walls sculpted by nature into intricate patterns of blues and whites, reflecting the ethereal light from an opening ahead. The crunch of ice underfoot and the drip-drip of melting water create a serene, echoing soundscape. As the camera moves closer, a gentle, ambient melody begins, swelling with the light from the cave's exit. The camera emerges from the narrow opening into a vast, sun-drenched valley, revealing a group of polar bears playfully sliding down an ice slope, their roars echoing with joy.

Prompt: A breathtaking, painterly 2D animated continuous visual narrative, rendered with the lush, vibrant, and slightly surreal, almost dreamlike, infused with the intricate, delicate detail of traditional Japanese woodblock prints (Ukiyo-e), follows a young, adventurous, and kind-hearted girl (perhaps with bright, curious eyes and wearing simple, practical, beautifully patterned traditional Japanese farm attire) as she befriends a colossal, gentle, ancient Forest Spirit. The Spirit is a magnificent, awe-inspiring creature, its form a harmonious blend of animal and plant – perhaps with moss-covered, antler-like branches, fur like shimmering leaves that change color with its mood, and eyes like deep, tranquil forest pools. They meet in a sun-dappled, sacred grove deep within an ancient, primeval forest, where impossibly tall, gnarled trees form a living cathedral and tiny, glowing, friendly forest sprites (Kodama-like) peek from behind mossy rocks and giant, fantastical mushrooms. The girl, initially awestruck, offers the massive Spirit a small, carefully cultivated offering – perhaps a perfectly ripe persimmon or a handful of wild berries – her gesture one of pure, innocent respect and affection. The Forest Spirit responds with a slow, gentle inclination of its massive head, its leafy fur rustling like a thousand whispers, and perhaps causes a shower of magical, luminous flower petals to drift down from the canopy, or a tiny, new sapling to sprout at the girl's feet. The animation captures the incredible, detailed textures of the forest, the Spirit's majestic yet gentle presence, and the profound, unspoken emotional connection forming between the child and this ancient guardian of nature. The color palette is a rich symphony of deep forest greens, earthy browns, vibrant floral hues, and the soft, magical glow of the sprites and the Spirit's own subtle luminescence. This continuous, sweeping visual journey is a celebration of the profound, often mystical, bond between humanity and nature, the innocence and courage of childhood, and the power of kindness and respect to bridge even the most fantastical of divides, an affectionate, visually intoxicating ode to ecological harmony and interspecies understanding. The only implied sounds are the gentle rustling of leaves, the distant calls of unseen forest birds, the girl's soft, respectful breathing, the Spirit's deep, resonant, almost inaudible hum, and a soaring, emotionally resonant, orchestral score.

Prompt: A handheld shot follows a wok as it’s expertly flicked, sending vibrant, sizzling vegetables tumbling over themselves in a flash of motion and steam. Audio: a metallic clank and a sharp whoosh.

Prompt: A keyboard whose keys are made of different types of candy. Typing makes sweet, crunchy sounds. Audio: Crunchy, sugary typing sounds, delighted giggles.

Prompt: The camera begins with a slow, elegant track along the richly paneled walls of a dimly lit, sophisticated hallway, the warm glow of the ornate wall sconces casting inviting reflections on the polished floor. Soft jazz music plays in the background. As we approach an arched entryway, the camera performs a graceful push-in, revealing a grand mirror and flickering candles, then smoothly pivots to the right, opening up to a luxurious home bar. The clinking of ice and the murmur of conversation become audible. The camera settles on a close-up of a perfectly crafted cocktail. "Welcome," a smooth, baritone voice says. "Care for a taste?" Suddenly, a renowned mixologist, known for his eccentric creations, steps into frame, followed by a playful, mischievous cat that jumps onto the bar, batting at a cocktail stirrer.

Prompt: A snow-covered plain of iridescent moon-dust under twilight skies. Thirty-foot crystalline flowers bloom, refracting light into slow-moving rainbows. A fur-cloaked figure walks between these colossal blossoms, leaving the only footprints in untouched dust.

Prompt: A woman, classical violinist with intense focus plays a complex, rapid passage from a Vivaldi concerto in an ornate, sunlit baroque hall during a rehearsal. Their bow dances across the strings with virtuosic speed and precision. Audio: Bright, virtuosic violin playing, resonant acoustics of the hall, distant footsteps of crew, conductor's occasional soft count-in (muffled), rustling sheet music.

Prompt: A close up in a smooth, slow pan focuses intently on diced onions hitting a scorching hot pan, instantly creating a dramatic sizzle. Audio: distinct sizzle.

Our partnership with Darren Aronofsky’s Primordial Soup

Darren Aronofsky’s Primordial Soup has partnered with Google DeepMind to explore AI as a tool to unlock the next chapter of human creativity.

Watch the film

New capabilities

Greater control, consistency, and creativity than ever before.

Add ingredients to your video

Make sure videos align with your creative vision by giving Veo reference images of a scene, a character, or an object to guide its generation. Now includes audio.

Input image

Input image

Input image

Prompt: Close up shot of woman with sunglasses on top of her head, gold hood earrings, is walking in the interior, she is lost and asks where everyone is and what's going on.

Match your style

Capture your desired aesthetic by providing a style reference image, and Veo will generate videos with the same visual style, from paintings to cinematic looks.

Input image

Output video

Prompt: Rendered in an intricate origami art style using complex, angular folds and crisp creases. A multi-layered diorama depicts a cute neighborhood street entirely from folded paper – houses with sharp rooflines, precise white picket fences, and layered, geometric flowers and rose bushes in vibrant paper hues. Focused lighting enhances the dimensionality. A vibrant origami cat, its body segmented by distinct, sharp folds, moves with articulated, deliberate steps along the paper sidewalk. Its limbs shift segment by segment, maintaining crisp creases as it progresses. The viewpoint tracks smoothly alongside the cat, revealing successive layers of the detailed papercraft neighborhood scrolling past, enhancing the scene's geometric depth and dimensionality.

Output video

Prompt: Rendered in an intricate origami art style using complex, angular folds and crisp creases within a detailed, multi-layered paper diorama featuring a sharply folded bus stop sign. Focused lighting enhances the geometric shapes. Five distinct origami children, constructed with precise folds defining summer clothes and angular backpacks, populate the scene. Two figures stand facing each other, their paper heads tilting slightly back and forth on sharp neck creases in articulated movements suggesting conversation. The remaining three figures execute a game: their folded leg sections bend sharply at distinct knee creases, then straighten abruptly, causing their entire forms to lift momentarily off the paper ground plane before settling back, repeating this crisp, angular jumping motion. Each movement is segmented and deliberate.

Output video

Prompt: Rendered in an intricate origami art style using complex, angular folds and crisp creases within a multi-layered paper diorama. Focused lighting enhances geometric shapes and dimensionality. A vibrant yellow school bus, constructed with sharp, precise folds defining its iconic shape, moves with deliberate, segmented progression along a winding road represented by a crisply folded paper strip. As the bus navigates the road's angular turns, its distinct paper facets catch and reflect the focused light, showcasing its geometric form. Its angular wheels might rotate sectionally or simply slide along the paper path. Above, the sky is a flat blue paper layer, featuring sharply folded, geometric white clouds and a bright, faceted origami sun casting crisp shadows across the layered scene.

Keep your characters consistent

Ensure characters maintain their appearance across different scenes in your videos by giving Veo reference images of your character.

Input image

Prompt: a monster walking toward the camera

Output video

Prompt: a cute monster dancing

Output video

Prompt: a cute monster swimming underwater

Output video

Prompt: a cute monster walking in a candy wonderland

Extend your scene

Extend clips into longer, more dynamic videos. Use the last second of your first shot to continue the story – while maintaining visual and audio consistency.

Input video

Prompt 1: Graceful dancer is slowly dancing to classical music.

Prompt 2: A male dancer comes in, gracefully dancing with the woman as classical music plays.

Prompt 3: More dancers show up on the stage.

Prompt 4: The classical music continues, and the dancers continue to dance.

Camera controls

Precisely control the framing and exact movement of shots in your video using camera controls.

Move back

Zoom in

Move up

Move right

First & last frame

Create smooth, artful, and epic transitions between images provided for the first and last frame.

First frame

Last frame

Prompt: Camera soars through the barn to the cowboy riding away.

Outpainting

Go beyond the original frame. Outpainting expands your video with new, matching parts that look real, helping it fit any screen size or shape.

Input video

Output video

Add object

Reimagine videos by introducing new objects - from realistic details to fantastical elements. Veo considers scale, interactions, and shadows to create a natural, realistic-looking video.

Input video

Output video

Prompt: Add a man with a torch

Remove object

Seamlessly eliminate unwanted objects from videos - from distracting details to large items. Veo preserves the scene's natural composition, interactions, and shadows.

Input video

Output video

Prompt: Remove spaceship

Character controls

Bring characters to life, using your body, face and voice to animate them.

Input video

Input image

Output video

Use your body to drive lifelike character movement and expressive actions that respond to your movements.

Motion controls

Define the exact movement of objects in your video. Select an object and define their path, and Veo will bring them to life in motion.

Flow

Built with creatives, for creatives. Flow enables you to create seamless cinematic clips, scenes, and stories using our most capable generative AI models.

Create with Flow

Benchmarks

Veo 3.1 is a new era for video generation. It's state of the art in text-to-video, image-to-video, text-to-audio+video generation, and realistic physics.

View benchmarks

Safety & Responsibility

We built Veo with responsibility and safety in mind. We block harmful requests and results, we test how new features might affect safety, and we have both our own teams and outside experts try to find and fix potential problems before release.

It's crucial to introduce technologies such as Veo in a responsible way. To achieve this, videos made with Veo will be marked with SynthID, our advanced technology for watermarking and detecting content generated by AI. Additionally, Veo outputs will undergo safety evaluations and checks for memorized content to reduce potential issues related to privacy, copyright infringement, and bias.

Limitations

While Veo continues to make incredible strides in video generation, creating videos with natural and consistent spoken audio, particularly for shorter speech segments, remains an area of active development. We're continuously working to refine audio synchronization and eliminate instances of incoherent speech.

Empowering production workflows

Discover how developers and studios are leveraging Veo to transform storytelling and production.

Try Veo

Flow

Gemini app

Google AI Studio

Gemini API

Google Vids

Vertex AI Studio

Acknowledgements

Veo 3 was made possible by key research and engineering contributions from Abhishek Sharma, Ágoston Weisz, Alina Kuznetsova, Ali Razavi, Aleksander Bulski, Aleksander Holynski, Ankush Gupta, Austin Waters, Ben Poole, Daniel Tanis, Derek Gasaway, Dumitru Erhan, Enric Corona, Evgeny Sluzhaev, Frank Belletti, Gabe Barth-Maron, Hakan Erdogan, Henna Nandwani, Hernan Moraldo, Ilya Figotin, Igor Saprykin, Jason Baldridge, Jeff Donahue, Jiawei Xia, Jimmy Shi, Khyatti Gupta, Kristina Greller, Kuang-Huei Lee, Kurtis David, Lizao (Larry) Li, Lijun Yu, Luis C. Cobo, Mai Gimenez, Medhini Narasimhan, Miaosen Wang, Mingda Zhang, Mohammad Babaeizadeh, Mukul Bhutani, Nikhil Khadke, Nilpa Jha, Oscar Akerlund, Pieter-Jan Kindermans, Poorva Rane, Rachel Hornung, Ricky Wong, Ruben Villegas, Ruiqi Gao, Ryan Poplin, Salah Zaiem, Sarah Xu, Sayna Ebrahimi, Scott Wisdom, Shlomi Fruchter, Sophia Sanchez, Tingbo Hou, Vikas Verma, Viral Carpenter, Xinchen Yan, Xinyu Wang, Yiwen Luo, Yukun Ma, Yukun Zhu, Zhichao Yin, and Zu Kim. All the clips were generated directly with Veo without modifications by Eleni Shaw, Signe Nørly, Andeep Toor, Gregory Shaw, Anne Menini, Matthieu Kim Lorrain, and Irina Blok.

We extend our gratitude to Ahmed Chowdhury, Andrew Audibert, Andrew Bunner, Andrew Pierson, Aparna Joshi, Asya Fadeeva, Austin Tarango, Bao Thach, Bihao Zhang, Bilva Chandra, Bogdan Damoc, Bryce Petrini, Cai Xu, Calin Cruceru, Chengrun Yang, Dana Kurniawan, David Reid, Emanuele Bugliarello, Ganesh GS, Gladys Tyen, Giorgos Vernikos, Greta Kintzley, Hakim Sidahmed, Hamid Mohammadi, Hiresh Gupta, Hiroki Furuta, Hongliang Fei, Huisheng Wang, Hui Zheng, Isa Liang, Izzeddin Gur, Jian Li, Jingjing Zhou, Jordi Pont-Tuset, José Lezama, Kangfu Mei, Karthik Narasimhan, Keyang Xu, Kory Mathewson, Larry Li, Lluis Castrejon, Liangke Gui, Mahyar Bordbar, Marek Sedlacek, Mikhail Dektiarev, Mitchell McIntire, Nick Pezzotti, Nick Tombari, Nitesh Bharadwaj Gundavarapu, Orly Liba, Pankil Botadra, Piyush Kumar, Ramin Mehran, Robert Geirhos, Sander Dieleman, Sirui Xie, Sherry Yang, Shubham Nauriyal, Shuo Han, Soňa Mokrá, Tamoghna Saha, Tim Salimans, Tom Hume, Woohyun Han, Xingyu Federico Xu, Xinyu Wang, Yelin Kim, Yong Cheng, Yuchi Liu, Yuexiang Zhai, Yutian Chen, Zerong Xi, Zhenkai Zhu, Zhisheng Xiao, and Zoltan Egyed for their invaluable partnership in developing and refining key components of this project.

Veo 2 controls were made possible by Abhishek Sharma, Aleksander Hołyński, Alina Kuznetsova, Andrew Marmon, Andrew Xue, Andrey Voynov, Anthony Mejia, Asaf Shul, Ben Poole, Brendan Shillingford, Dawid Górny, Dina Bashkirova, Dmitry Lagun, Emanuele Bugliarello, Enric Corona, Henna Nandwani, Inbar Mosseri, Istvan Hernadvolgyi, Jess Gallegos, Jieru Hu, Kristina Greller, Luciano Sbaiz, Matan Cohen, Miaosen Wang, Mingda Zhang, Nikos Kolotouros, Nick Pezzotti, Philipp Henzler, Ricky Wong, Roni Paiss, Rui Huang, Ruiqi Gao, Ryan Webb, Serena Zhang, Shiran Zada, Siyang Li, Tali Dekel, Tatiana López, Tayniat Khan, Thomas Kipf, Tingbo Hou, Tobias Pfaff, Tom Murray, Xin Yuan, Xinyu Wang, Yulia Rubanova, Yusuf Aytar, and Zhichao Yin.

We extend our gratitude to Alex Rav Acha, Amir Hertz, Andrew Pierson, Ankush Gupta, Anthony Tripaldi, Austin Tarango, Ben Bariach, Bilva Chandra, Budianto Budianto, Carl Doersch, Changchang Wu, David Minnen, David Yao, Dexter Allen, Dilara Gokay, Dumitru Erhan, Eric Lau, Erik Gross, Florian Schroff, Frank Belletti, Gitartha Goswami, Hang Qi, Hao Wang, Hao Zhou, Harsimran Kaur, Itzhak Garbuz, Jason Zhang, Jenny Brennan, Jessica Seah, Jiaping Zhao, Jordi Serrano Berbel, Kan Chen, Ke Yu, Kory Mathewson, Kurtis David, Lluis Castrejon, Luis C. Cobo, Mahyar Bordbar, Manika Puri, Matthew Burruss, Matthew Levine, Matthieu Kim Lorrain, Medhini Narasimhan, Metin Toksoz-Exley, Michael Chang, Michael Milne, Navin Sarma, Nick Matarese, Noah Snavely, Pankil Botadra, Pieter-Jan Kindermans, Reggie Ballesteros, Richard Tucker, Ryan Poplin, Sasha Brown, Shantanu Bhattacharya, Siavash Khodadadeh, Soumyadip Ghosh, Srimon Chatterjee, Ting Liu, Tom Hume, Troy Chinen, Vika Koriakin, Viral Carpenter, Xiang Li, Xuemei Zhao, Xuhui Jia, Yael Pritch, Yedid Hoshen, Yi Yang, Yuan Zhong, and Yutian Chen.

Special thanks to Douglas Eck, Aäron van den Oord, Eli Collins, Koray Kavukcuoglu, Demis Hassabis and Sergey Brin for their insightful guidance and support throughout the research process.

We also acknowledge our infrastructure partners Abhinash Giri, Allen Wu, Jon Blanton, Praseem Banzal, Ricky Liang, and Shariar “Nafi” Rouf. And the many other individuals who contributed across Google DeepMind and our partners at Google.