Gemma
Google's family of small, capable open-weight models built from Gemini research.
Gemma is Google's line of open-weight models, derived from the same research as Gemini but small enough to run on a single GPU or a laptop. The family includes variants tuned for specific tasks and sizes.
It is a strong choice when you want a compact, well-supported open model with a major lab behind it.
Where it fits
A fit for self-hosted use cases that need a small, efficient model with the polish of a major lab.
Where it does not fit
Small models trade away peak capability: not the choice for the hardest reasoning tasks.
Alternatives to Gemma
The closest options on the map, and where each fits instead.
-
Llama Open-weight model family
The default open-weight choice when models must run inside your own environment and you want the widest tooling support.
Unlike Gemma Maturity: mature
-
Phi Open-weight model family
A fit when you need genuine reasoning from a small, cheap-to-run model with a permissive licence.
-
Qwen Open-weight model family
A strong fit when you need to size a self-hosted model to specific hardware, or when multilingual coverage matters.
Unlike Gemma Hosted API
-
Mistral (open models) Open-weight model family
A fit when you want capable, efficient open weights with a permissive licence that is straightforward for commercial use.
Pairs with
Commonly used alongside it.
