Overview
Gemma 3 is Google's latest open-weight model family, available on Ollama for local deployment. Built from the same research as Gemini, Gemma offers strong performance in compact sizes optimized for consumer hardware.
Sizes
2B / 4B / 12B / 27B
Context
32K-128K tokens
VRAM
2-20GB
License
Gemma Terms
✅Strengths
- ✓Google research quality, open weights
- ✓Excellent performance per parameter
- ✓Strong multilingual support
- ✓Good code generation capabilities
- ✓Well-optimized for Ollama
⚠️Weaknesses
- ✗License restrictions on commercial use
- ✗Smaller context than Llama 3.2
- ✗Requires local hardware
- ✗Less community support than Llama
Best Use Cases
💻 Local Coding
IDE assistance
🌐 Translation
Multilingual apps
📝 Writing
Content creation
🔬 Research
Analysis, summaries
📊 Classification
Sentiment, topics
🎓 Education
Learning AI
Ollama Command
ollama run gemma3:12b
Available variants: gemma3:2b, gemma3:4b, gemma3:12b, gemma3:27b