Skip to content

Pick your style: Denkbaar

Book a conversation

Articles

Knowledge

What is Gemma?

Gemma is Google's family of open models that you can download and run yourself.

An employee places a compute card in a local server.
You can run Gemma on your own equipment. Illustration: Denkbaar.

Gemma is a family of open models from Google. You can download the weights and run them on your own machine or server.

What you read about it

  • On 21 February 2024, Gemma appears in two sizes, 2B and 7B, from the same research as Gemini: Google releases Gemma.
  • Gemma 2 follows on 27 June 2024 in 9 and 27 billion parameters. The largest model runs on one data-centre GPU: Gemma 2.
  • Gemma 3 arrives on 12 March 2025 and focuses on models that run on one GPU or TPU: Gemma 3.
Three modules of different sizes stand beside a compute card.
Gemma appeared in different models and sizes. Illustration: Denkbaar.

Why it matters to you

A model that fits on one compute card can run at a European host or on your own server. Always read the licence: it sets what you may do with it in business. When you get a quote, also ask who does updates and fixes faults.

Two people examine a document beside a server.
Check the licence and agree who handles maintenance. Illustration: Denkbaar.

Written on 13 November 2026, based on the articles and sources below.

Sources

  1. Google: Gemma: Google introduces new state-of-the-art open models
  2. Google launches Gemma 2, its next generation of open models (Google)
  3. Gemma 3

Topics: open-sourcegoogle