curatedai.net
Directory Categories Radar Models Leaderboard Guides Prompts Skills Briefing Newsletter
Light Dark
← Back
← All Gemma models
LLMS • CURATED • UPDATED FEB 5, 2026

Gemma

Google's open-source lightweight LLM

Gemma is Google DeepMind's family of open-source large language models, serving as lightweight versions of Gemini. Available models include Gemma 1 (February 2024), Gemma 2 (June 2024), and Gemma 3 (March 2026) with variants like PaliGemma for vision-language tasks and MedGemma for medical applications. Available in multiple sizes (2B, 7B, and larger variants). Designed for research, education, and commercial applications with permissive licensing. Trained on similar data and methods as Gemini models but optimized for open-source deployment. Available through Hugging Face, Kaggle, and Google Cloud Vertex AI.

QUICK FACTS

Pricing Free
Platforms api, local
API Yes
Open Source Yes
Modalities LLMs
Best For Best for Research
Date Added 2026-02-05

WHY WE PICKED IT

Google's open-source LLM family with strong performance, permissive licensing, and specialized variants for vision and medical applications.

STRENGTHS

LIMITATIONS

GETTING STARTED

Visit Google's Gemma documentation at ai.google.dev/gemma or deepmind.google/technologies/gemma. Download model weights from Hugging Face (huggingface.co/google) or Kaggle (kaggle.com/models/google/gemma). For API access, use Google Cloud Vertex AI. Deploy locally using PyTorch, JAX, or TensorFlow. Gemma 3 (March 2026) is the latest version with improved performance. Follow Google's documentation for setup, fine-tuning, and deployment. Free to use for research and commercial applications under permissive licensing.

QUICK TIPS

1 Choose appropriate model size based on your needs
2 Use PaliGemma for vision-language tasks
3 Leverage MedGemma for medical and healthcare applications
4 Take advantage of permissive licensing for commercial use
5 Explore fine-tuning for domain-specific applications

OFFICIAL LINKS

Website↗ Documentation↗ Hugging Face↗ Kaggle↗ GitHub↗

SIMILAR TOOLS

Claude Fable 5 NotebookLM Claude Opus 5 GPT-5.6 Sol Kimi K3

USE CASE EXAMPLES

Research and Education

Use Gemma for AI research, education, and academic projects with full access to model weights and training data.

STEPS:
  1. Download Gemma models from Hugging Face or Kaggle
  2. Set up research environment with PyTorch or JAX
  3. Conduct experiments and fine-tuning
  4. Publish research with full transparency

Vision-Language Applications

Use PaliGemma variant for tasks requiring both image and text understanding.

STEPS:
  1. Download PaliGemma model weights
  2. Set up vision-language pipeline
  3. Process images and text together
  4. Generate multimodal outputs

PRICING

Free Completely free
📚

FEATURED IN GUIDES

Open Models: What They Cannot Do

The benchmark gap has largely closed: open models now score within a point of the closed frontier on...

Hardware For Local AI Models: What The Memory Shortage Changed

Capacity decides whether a model runs. Bandwidth decides how fast it talks, and it is the number mos...

How To Run AI Models Locally: Hardware, Quantisation And Engines

Running a frontier open model locally is a memory problem before it is a speed problem, and a bandwi...

Uncensored AI Models: What Abliteration Actually Does To Them

Abliteration removes a model's ability to refuse by deleting one direction from its activations. It ...

Which LLMs Actually Deliver in 2026?

Comprehensive comparison of the best large language models in 2026 including ChatGPT, Claude, Gemini...

EXPLORE ALTERNATIVES

View Gemma Alternatives (2026) →

Compare Gemma with 5+ similar llms AI tools.

❓

FREQUENTLY ASKED QUESTIONS

Q

Is Gemma free?

A

Yes, Gemma is completely free to use.

Q

Does Gemma have an API?

A

Yes, Gemma offers an API for programmatic integration.

Q

What is Gemma best for?

A

Gemma is best for Research. Gemma is Google DeepMind's family of open-source large language models, serving as lightweight versions of Gemini. Google's open-source LLM family with strong performance, permissive licensing, and specialized variants for vision and medical applications.

Q

What platforms does Gemma support?

A

Gemma supports api, local.

Q

Is Gemma open source?

A

Yes, Gemma is open source. You can access the source code on GitHub at https://github.com/google/gemma_pytorch.

Q

How do I get started with Gemma?

A

Visit Google's Gemma documentation at ai.google.dev/gemma or deepmind.google/technologies/gemma. Download model weights from Hugging Face (huggingface.co/google) or Kaggle (kaggle.com/models/google/gemma). For API access, use Google Cloud Vertex AI. Deploy locally using PyTorch, JAX, or TensorFlow. ...

Q

How do I use Gemma?

A

Gemma is a large language model for text generation, analysis, and conversation. Use the API for programmatic access. Enter prompts or questions to get responses. It excels at open-source with permissive licensing.

🏷️

FEATURED ON CURATEDAI

Work on Gemma? You're hand-reviewed in our directory. Add this badge to your site — it links back to this profile.

Featured on CuratedAI