Google has announced the launch of its latest “most capable” large language model, trained using its custom-designed AI accelerators, Cloud TPU v4 and v5e. Built from the ground up to be multimodal, Gemini is Google’s long-awaited answer to ChatGPT. It can generalize and seamlessly understand, operate across, and combine different types of information including text, images, audio, video, and code. In addition to being natively multimodal (empowering users to transform any type of input into any type of output), Gemini also features reasoning and advanced coding capabilities.

“Our first version, Gemini 1.0, is optimized for different sizes: Ultra, Pro and Nano. These are the first models of the Gemini era and the first realization of the vision we had when we formed Google DeepMind earlier this year. This new era of models represents one of the biggest science and engineering efforts we’ve undertaken as a company,” says Google CEO Sundar Pichai.

Starting today, Google Bard (the company’s generative AI chatbot) will use a specifically tuned version of Gemini Pro in English for more advanced reasoning, planning, and more. Early next year, the company plans to introduce Bard Advanced to provide users with access to its most advanced models and capabilities — starting with Gemini Ultra. Also, Pixel 8 Pro is the first smartphone engineered to run Gemini Nano; it will begin rolling out to power new features including Summarize in the Recorder app and a developer preview of Smart Reply in Gboard, available to try first with WhatsApp and coming to more messaging apps next year.

Android developers can sign up for our Early Access Program to build with Gemini Nano via Android AICore. On December 13, Gemini will be available in Vertex AI and AI Studio.

“I’m genuinely excited for what’s ahead, and for the opportunities Gemini will unlock for people everywhere,” Pichai adds.