Members of the Gemma model family share a small, lightweight design but specialize in different tasks. Choosing within the family means matching the model to the work, from general text to code, safety, vision, and medical data:
- Gemma — a small, lightweight open model supporting text generation, summarization, and extraction.
- Gemma 2 — an open model supporting text generation, summarization, and extraction.
- ShieldGemma 2 — instruction-tuned models for evaluating text and image safety against defined policies.
- PaliGemma — an open vision-language model combining SigLIP and Gemma.
- CodeGemma — a lightweight open model for coding tasks, including code completion, generation, and understanding.
- TxGemma — generates predictions, classifications, or text based on therapeutic-related data, for building AI models with less data and compute.
- MedGemma — a collection of Gemma 3 variants trained for performance on medical text and image comprehension.
- MedSigLIP — a SigLIP variant trained to encode medical images and text into a common embedding space.
- T5Gemma — a family of lightweight encoder-decoder research models.