Deep Learning
Gemini Security Improvements
Adjusting ratings for adaptive attacksBasic mitigations have shown promise against basic, non-adaptive attacks, significantly reducing attack success rates. However, malicious actors are...
Deep Learning
Advanced dialog and sound generation in Gemini 2.5
Safety and responsibilityWe actively assessed potential risks at every stage of the development process for these native audio features, using what we learned to...
Deep Learning
How we support better tropical cyclone prediction with artificial intelligence
Weather Lab users can view and compare forecasts from various artificial intelligence and physics-based models. Together, these forecasts can aid weather agencies...
Deep Learning
We are expanding our Gemini 2.5 family
We designed Gemini 2.5 as a family of hybrid reasoning models that deliver incredible performance while being at the highest level Pareto frontier costs...
Deep Learning
Gemini 2.5: Updates to our thinking model family
Today we're excited to share updates to the entire Gemini 2.5 family:Gemini 2.5 Pro is generally available and stable (no...
Deep Learning
How Gemma’s model helped discover a recent potential path for cancer therapy
We gave our recent ones C2S 27B scale model the task: Find a drug that works like conditioned reinforcerone that will enhance the immune...
Deep Learning
Introducing Veo 3.1 and advanced Flow capabilities
Five months ago we presented Flowour Veo-powered AI video creation tool, and inspired us with the creativity it has ignited with over 275 million...
Deep Learning
We present the Gemini 2.5 computer support model
Earlier this year we mentioned that we provide developers with desktop capabilities through the Gemini API. Today we publish Gemini 2.5 computer usage modelour...