Gemini is a generative artificial intelligence based on a language model developed by the company Google. It was previously known as Bard. Gemini competes for user popularity, especially with Chat GPT.

Gemini's sub-programs include the following services:

• Gemini 2.5 Flash, which offers fast performance for daily tasks.

• Gemini 2.5 Pro. The Gemini version released in March 2025, which is capable of demanding task resolution.

• Canvas (a collaborative mode), where one can create, format, and refine documents or code. For example, you can highlight a certain string of characters in a document, after which you can give Canvas a command to improve the text.

• Guided Learning, which functions as a personal learning tutor.
• Gems, which are customized versions of Gemini that provide tailored responses. 
• Gemini Deep Research, which performs comprehensive analyses on given topics. Usage is limited to five reports per user within a 30-day period.  Note! Deep Research is accessed by selecting the 2.5 Flash mode in the Gemini user interface.

Gemini sub-programs can be accessed in the Gemini user interface. The image generation tool Nano Banana can also be used via Google AI Studio. 

Note! The above sub-applications are available intermittently in the version used with a Metropolia ID. Sometimes they are offered, and sometimes they are not.

Address

• https://gemini.google.com

Does Gemini use prompts to train its AI models?

According to its own statement, Google does not use prompts to train its models if Gemini is used through Google Workspace, which in practice means with a Metropolia User ID. Conversations and inputs are not reviewed by Google employees.

Gemini is subject to the same privacy and security policies as other core Google services used with a Metropolia account, such as Google Drive.

In Finnish