Gemini is Google DeepMind’s family of natively multimodal AI models, built to work across text, images, audio, and more. Google DeepMind develops the family, which is offered through Google’s developer APIs and consumer products.
As of June 2026, the Gemini technologies page presents a lineup organized around Pro, Flash, and specialized variants. It names Gemini 3.5 Flash and Gemini 3.1 Pro, a Gemini 3.1 Deep Think reasoning variant, and a lightweight Gemini 3.1 Flash-Lite, with Gemini 3.5 Pro listed as “coming soon.” The page also references Gemini-based models for other modalities, including Gemini Omni, Gemini Audio, Gemini Robotics, a Gemini Embedding model, and a Gemini Image model. The page does not state explicit calendar release dates, and version numbers shift quickly.
Distribution is primarily through Google’s APIs and applications.
Why business readers should care: Gemini is the AI line embedded across Google’s products and cloud, so its current tiers signal what will reach hundreds of millions of users through tools many organizations already pay for.