Google DeepMind Gemini

Google DeepMind is the artificial intelligence research division formed from the merger of Google's DeepMind and the Brain team, working on advancing machine learning and building general-purpose AI systems. This page presents the Gemini family of large multimodal models, which can process and reason across text, images, audio, video, and code. Gemini is integrated across Google products including Search, Android, and Workspace, and is available to developers through Google's AI development platforms and APIs. The models come in multiple sizes optimized for different tasks, from on-device efficiency to large-scale reasoning. Google DeepMind publishes research on reinforcement learning, protein folding through AlphaFold, and safety alignment, and offers documentation, benchmarks, and interactive demos for researchers and builders exploring the capabilities of these models.