Gemini is an AI assistant made by Google. It is an interface to a multimodal large language model (LLM): a neural network that reads and generates text, audio, images and more. Google launched it in March 2023 under the name Bard. It runs on the web and in mobile apps; its newest model is Gemini 4 Argon.
A person writes or speaks a request in a conversation, and Gemini answers: it summarizes uploaded documents, helps with code, drafts text, and generates images and videos. Gemini Live is a spoken conversation. Deep Research is an agent feature: it turns a prompt into a research plan, browses up to hundreds of websites (and, with permission, the user's Gmail, Drive and Chat), and writes a multi-page report in minutes. Gemini also works inside Google apps such as Gmail and Docs, and in the Chrome browser.
The app is free with a Google account, with usage limits. Google AI Plus, Pro and Ultra subscriptions raise the limits and add newer models; Gemini Enterprise is priced per seat for businesses. Developers reach the models through the separate Gemini API.