Gemma
Google open models for local and responsible AI builds
Gemma is Google's open-weight sibling to Gemini. It gives developers smaller, tunable models that can run locally, on a single accelerator, or in controlled production environments. Gemma matters because many real deployments do not need a giant frontier model; they need a model that is transparent enough to adapt, cheap enough to run, and responsible enough to evaluate.
Variants
5
Best at
open weights with Google tooling
Max context
128K
tokens
Pricing
Open weights
$0
per self-hosted model weights
Google AI / Vertex
varies
per hosted deployment
Variants
Sort the table by context window or cost to find the right variant. Click any version below for a battle card with ranks, pricing notes, and official links.
| Modalities | ||||
|---|---|---|---|---|
Gemma 3 27B gemma-3-27b-it | 128K | varies / varies | 2025 | textvision |
Gemma 3 12B gemma-3-12b-it | 128K | varies / varies | 2025 | textvision |
Gemma 3n E4B gemma-3n-e4b | 32K | varies / varies | 2025 | textvisionaudio |
Gemma 3n E2B gemma-3n-e2b | 32K | varies / varies | 2025 | textvisionaudio |
CodeGemma codegemma | 8K | varies / varies | 2024 | textcode |
Battle card
Context rank
#2
within Gemma
Capability rank
#1
modalities + reasoning
Weights
Open
self-hostable if licensed
Best fights to pick
- mobile assistants
- offline demos
- accessibility workflows
- edge multimodal apps
Rankings are Tendril directory ranks, computed from the model data shown here. Public benchmark leaderboards change often, so official docs and current benchmark pages should be checked before buying or deploying.
Learn
Lessons about this model
Structured lessons that cover Gemma directly or put it in context alongside its rivals.
Gemma: Family Map
Gemma: Flagship Picker
Gemma: Cost And Context
Gemma: Reasoning Mode
Gemma: Multimodal Workflows
Gemma: Coding And Software Tasks
Gemma: Agents And Tool Use
Gemma: Local Or Private Deployment
Gemma: Benchmarks And Reality Checks
Gemma: Safety And Limits
Gemma: Prompt Recipes
Gemma: Industry Use Cases
Gemma: Compare Against Rivals
Gemma: Build A Mini Workflow
Gemma: Capstone Selection Memo
Check yourself
Quizzes
Short, mixed-difficulty quiz sets on Gemma and its model family.
Gemma Model Family Quiz
5 questions
A short quiz on where Gemma fits, which variants matter, and how to choose it responsibly.
Start quiz →Hands-on
Try these prompts
Ready-made prompts that show Gemma at its best. Use them in your own AI workspace, then compare the output with what you learned in Tendril.
Gemma same-task bakeoff
BuildersCompare Gemma against a cheaper and stronger rival on one real task.
Use Gemma for this task, then compare with one cheaper model and one stronger model: summarize a 10-page policy document into risks, decisions, owners, and open questions.
Gemma failure-mode log
CreatorsTrack where Gemma fails instead of only saving the best output.
Run this task three times. For every answer, mark failures as factual error, formatting error, weak reasoning, refusal, or missing citation.
Print & keep
Printable reference
One-page summaries and flowcharts — great for desks, classrooms, or study sessions.
Go deeper
Official resources
Straight from the lab — docs, API references, and the chat surfaces you can try today.
Strengths
- open weights with Google tooling
- strong local deployment story
- multimodal small-model options
Limits
- less capable than Gemini frontier models
- hardware constraints still matter at 27B
- some variants have shorter context