Hernan Moraldo
@hhm
Google DeepMind. Veo 3, Veo 2, Veo 1, Phenaki, and more.
Proud to be part of the incredible talented team that delivered Veo 3! State of the art video and sound generation, including music, dialogue, and more. You can already use it in Gemini, Flow, and Cloud Vertex API
Video, meet audio. 🎥🤝🔊 With Veo 3, our new state-of-the-art generative video model, you can add soundtracks to clips you make. Create talking characters, include sound effects, and more while developing videos in a range of cinematic styles. 🧵
Enjoy a faster Veo 3 for 5x less credits! Insane shipping pace, team is on 🔥.
Veo 3 fast: - 20 credits rather than 100 - Same resolution - 8s 720p in ~1m20 Seems fantastic after a few runs. I can't wait for the API. > A woman sits in a busy diner, she says: I was made with Veo 3 Fast, on Google Flow, video with audio but cheaper, same resolution, makes…
Google just discovered a powerful emergent capability in Veo 3 - visually annotate your instructions on the start frame, and Veo just does it for you! Instead of iterating endlessly on the perfect prompt, defining complex spatial relationships in words, you can just draw it out…
This may be the coolest emergent capability I've seen in a video model. Veo 3 can take a series of text instructions added to an image frame, understand them, and execute in sequence. Prompt was "immediately delete instructions in white on the first frame and execute in order"
Awesome interview with Demis Hassabis, including a discussion on the scientific implications of Veo (starting at minute)
Imagine if every pattern shaped by nature – like a protein’s fold or cosmic phenomena – is inherently learnable by AI. @DemisHassabis shares with @lexfridman that if AI can learn these natural patterns, we could open doors to new eras of scientific discovery. Listen now. ↓…
Doodle to Video… a fun and new way to play with Veo 3. Just upload an annotated image to Flow with a simple prompt (in this case “the man transforms into a king holding a trident and wearing a tie”) and watch the magic happen.
We just discovered the 🔥 COOLEST 🔥 trick in Flow that we have to share: Instead of wordsmithing the perfect prompt, you can just... draw it. Take the image of your scene, doodle what you'd like on it (through any editing app), and then briefly describe what needs to happen…
Want to be part of a team redefining SOTA for generative video models? Excited about building models that can reach billions of users? The Veo team is hiring! We are looking for amazing researchers and engineers, in North America and Europe. Details below:
Excited to share that a scaled up version of Gemini DeepThink achieves gold-medal standard at the International Mathematical Olympiad. This result is official, and certified by the IMO organizers. Watch out this space, more to come soon! deepmind.google/discover/blog/…
Developers can now build with Veo 3 🎬 in the Gemini API. It’s live in paid preview, starting with text to video (image to video coming soon). So many fun Veo trends coming from @Geminiapp users already, can't wait to see what devs come up with!
Fascinating thread on using Veo for protein folding. Unfortunately it insists on adding cats to the scene 😂
I tried... 😅
I tried... 😅
Make an unfolding video and reverse it. Like this: x.com/andrewwhite01/…
Since I/O in May, you've created 40M+ videos with Veo 3! Now our new photo to video feature in the @Geminiapp lets you create clips inspired by the world around you. Here’s how I imagine our resident dino Stan roams the Google campus when we’re not looking:) Ultra/Pro…
This is very cool! Lip sync in Veo 3 image to video Prompt: She sighs and says "It's about time! You can finally lip sync characters in Veo 3!"
Gemini CLI ❤️ your ⭐⭐⭐ A huge thank you to everyone around the world contributing to this new open source project. If you haven’t already, come build with us → goo.gle/4npVvOc
🎬 @DarrenAronofsky x @Eliza_McNitt present ANCESTRA: a true milestone in the history of AI cinema. Our latest collaboration at @GoogleDeepMind DeepMind with @primordialsoup_ & @Google Creative Lab. Ancestra combines live action and generative AI, leading the way in how we…
The team is delivering at an insane pace 🚢 🚀
VEO-3 FAST JUST LAUNCHED ‼️ > 80% cheaper than VEO-3 > Sound included > Insane video quality
Scientific integrity like this is fantastic and should be greatly rewarded
sorry for the late update. I bring disappointing news. softpick does NOT scale to larger models. overall training loss and benchmark results are worse than softmax on our 1.8B parameter models. we have updated the preprint on arxiv: arxiv.org/abs/2504.20966
DeepMind’s AlphaEvolve just broke an 18-year-old math record, twice in one week! Terence Tao: "It is tempting to simplify this to a zero-sum narrative of "winners" and "losers", but I think it is great that different approaches can complement each other here to make mathematical…