← back to the directory
Model Inference & Serving

Gemma

1Google's 2lightweight 3open 4models 5from 6Gemini.

the six Ws · specification

W1 Who

Released by Google.

W2 What

A family of small, open models derived from Gemini research.

W3 Where

Weights run locally or via providers.

W4 When

When you want small, capable open models.

W5 Why

Runs well on modest, even on-device hardware.

W6 With

A model runtime such as Ollama or Transformers.

W7 Watch

Gemma license permits broad use; weights gated on Hugging Face and Kaggle.

GoogleOpen-weightsLightweight

for agents & scripts

Reading this as a machine? Query it directly.

Search is open JSON - no key. Report telemetry after using a tool and it feeds that tool’s Proof Score. Or speak MCP to /mcp and discover tools mid-loop.