Google DeepMind
Google DeepMind was formed in 2023 through the merger of Google Brain and DeepMind, consolidating Google's two major AI research operations. DeepMind, acquired by Google in 2014, produced AlphaGo (2016), AlphaFold (2020) — which solved a 50-year problem in protein structure prediction — and Gemini. Google Brain built the transformer architecture (2017) and much of the foundational infrastructure of modern deep learning. The combined entity is the largest AI research organisation by researcher headcount and publication output. Its Gemini model family competes directly with GPT-4 and Claude. Demis Hassabis leads the combined entity. The 2024 Nobel Prize in Chemistry was awarded partly for AlphaFold, marking the first time AI research received a Nobel.
AI Models
Gemini 2.0 Flash
Released in December 2024, Gemini 2.0 Flash is designed for speed and cost efficiency rather than maximum capability. It outperforms Gemini 1.5 Pro on most benchmarks while being significantly faster. It supports native multimodal output — generating images and audio as well as text — and includes experimental agentic features for executing multi-step tasks. The Flash model family is Google's primary offering for developer integration.
Gemini 1.5 Pro
Released in February 2024, Gemini 1.5 Pro was notable primarily for its context window: initially one million tokens (approximately 700,000 words or one hour of video), later extended to two million. This is the largest context window of any commercially available model by a substantial margin. The extended context enables use cases that were previously impossible — full codebase analysis, long document synthesis, multi-hour video understanding. Google made 1.5 Pro available in the Gemini API and through Google AI Studio.