"Gemini" is Google's name for its main platform of large language models, reasoning models, and image/video creation tools. There are usually several available to users, whether they are on the free tier or a paid tier.
As of spring 2025, they include:
- 2.0 Flash: A large language model that can be used for many tasks, such as brainstorming, looking over papers for edits, and summarizing text.
- 2.5 Professional: The top-of-the-line Gemini model. It is a so-called "reasoning" model which gives much more in-depth answers and works through a series of steps to generate raw material for its answers.
- Imagen 3: Google's primary image generation model, accessible through the main chatbot.
- Veo2: Google's generative AI-powered video creator and editor, which is a standalone model.
Some of these are only available on the Gemini Advanced tier, or (like 2.5) are available on the free tier but have a limited number of uses per month. Google is currently running a promotion for college students, making the Advanced tier available for free through the 2025-2026 school year; details are available here.
Furthermore, there are a few ways to expand Gemini's powers within the chat:
- One can add files, including images, to provide Gemini data or context for your chat.
- "Canvas" opens a window within the chat that can execute code or show an editable file which Gemini can collaborate on.
- "Deep Research" is a mode that allows Gemini to execute a long series of search steps, and synthesizing many sources from the internet, to provide a report on a research question. Currently this will predominately draw on "ordinary" websites rather than academic sources, although it will occasionally find sources on open access scholarly websites.
