Google DeepMind

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. Veo 3 also has: ✅ Superior visual quality  ✅ Excellent real-world physics  ✅ Prompt following like never before Veo 3 is available today for Ultra subscribers in the United States in the Gemini App. Find out more about where you can use it ↓ https://goo.gle/4k5Uqck  #GoogleIO

We’re launching Gemini 2.5 Deep Think: an enhanced reasoning mode for Gemini that extends its "thinking time" to help tackle harder problems. We put it into the hands of mathematicians who explored what it can do. 🧠 💭 Deep Think first generates parallel streams of thought, before comparing, contrasting, and refining these ideas to arrive at better answers. 💡 Its strengths goes beyond the realms of math and science, performing well at work that needs creativity and strategic planning - including tackling tough coding problems as well as iterative web design. Deep Think is now rolling out to Google AI Ultra subscribers. Find out more → https://lnkd.in/euyuUPSR

Our new AI model AlphaEarth Foundations is mapping the planet in astonishing detail. 🌏🔍 The system’s public datasets will help scientists better plot and track planetary changes over time. This could include critical issues like water resources, deforestation, urban expansion, and more – in hours, instead of weeks or months. Organizations like the United Nations FAO, Group on Earth Observations and MapBiomas are already using it to create powerful custom maps and drive real-world insights. The annual embedding datasets from our model are available now in Google Earth Engine. AlphaEarth Foundations is a significant step forward in understanding our changing planet and we’re excited to see how it can accelerate work in vital mapping initiatives. Find out more → https://goo.gle/4l7fZsC

Our new state-of-the-art AI model Aeneas transforms how historians connect the past. 📜 Ancient inscriptions often lack context – it's like solving a puzzle with 90% of the pieces lost to time. Aeneas helps researchers interpret and situate inscriptions in their past context. By transforming each ancient text into a unique historical fingerprint, Aeneas can identify similarities across 176,000 Latin inscriptions. In our study, historians found these ‘parallels’ to be helpful research starting points 9 out of 10 times - improving their confidence by 44%. We tested Aeneas on the Res Gestae Divi Augusti – one of the most debated ancient inscriptions. Without prior knowledge, it successfully mapped out the leading scholarly theories on its dating, showing how AI can help model history in a quantitative way. 📊 Aeneas is now accessible through: 👉A website for researchers 🧑💻Open-source code and dataset 📚Syllabus for classrooms 🏛️Upgraded Ithaca ancient Greek model We’re excited to see how more people use this work to uncover the past. Find out more → https://goo.gle/4kVkh6n

Scale your production apps with the stable version of Gemini 2.5 Flash-Lite. ⚡ It’s faster than our 2.0 Flash models, more cost-efficient, and outperforms 2.0 Flash-Lite across coding, math, reasoning, and multimodal understanding. Start building → ai.studio

An advanced version of Gemini with Deep Think has officially achieved gold-medal performance at the International Mathematical Olympiad. 🥇 Here’s why this breakthrough matters ↓ 🧮 The IMO is the world’s most prestigious annual math competition, presenting six problems to participants involving: geometry, algebra, number theory, and combinatorics. 🔢 Gemini was given the same 6️⃣ problems and time limit - 4.5 hours - as human competitors, and still produced rigorous mathematical proofs. 🧠 With Deep Think, an enhanced reasoning mode, our model was able to simultaneously explore and combine multiple possible solutions before giving a definitive answer. 💭 We pushed this version further by giving it more thinking time, access to high-quality solutions to previous problems and pre-prepared hints and tips on how to find the best answers. These results show we're closer to building systems that can solve more complex mathematics. A version of this Deep Think model will soon be available to trusted testers, including mathematicians, before rolling out to Google AI Ultra subscribers. Find out more → https://lnkd.in/eFT6HaXr

Start building with Veo 3: our state-of-the-art video generation model now available in paid public preview via the Gemini API and Google AI Studio. 🎨 Here's how to try it → https://goo.gle/4eUk9m9

We’re bringing Gemini 2.5 Pro to AI Mode: giving you access to our most intelligent AI model, right in Google Search. 🔍 With advanced reasoning capabilities, watch how it can tackle incredibly difficult math problems, with links to learn more ↓

Presenting The Great Voyage: a 3-minute film made with AI, inspired by Victorian silent cinema. 🎬 Made by our in-house creative team, it uses: ✍️ Gemini for inspiration and prompt writing 🖌️ Imagen to generate starting visuals 🎥 Veo’s image-to-video capabilities to add motion and craft the narrative 🎵 Music AI Sandbox for a soundtrack Got any questions? Let us know in the comments. ⬇️

One of the projects I've been working on in collaboration with the amazing Project Zero folks is called "Big Sleep." It's an agentic AI system developed to hunt for critical security vulnerabilities. It is a prime example of how AI is transforming security. One of the really unexpected use-cases for me has been in threat intel. Recently, we were able to use AI to identify, reproduce and help patch a critical vulnerability based on partial intel which we believe is a world first. While we hope to share more technical details in the coming months, you can read more about Big Sleep, and other security updates from across Google, in our latest blog https://lnkd.in/gfZSTN25