Model family
Gemma Models
Gemma models are useful for efficient local, app, and multimodal workflows, with small-to-mid-size variants that are practical for developers.
Best for
Chat
Use this family hub to compare Gemma variants for chat workflows, then open the detail page for deeper deployment notes.
Local
Use this family hub to compare Gemma variants for local workflows, then open the detail page for deeper deployment notes.
Efficient
Use this family hub to compare Gemma variants for efficient workflows, then open the detail page for deeper deployment notes.
Multimodal
Use this family hub to compare Gemma variants for multimodal workflows, then open the detail page for deeper deployment notes.
Source box
This family currently includes 10 records tied to an exact published checkpoint and 2 broader or unresolved records. Identity is recorded per model so a representative checkpoint is never treated as the whole family.
Identity checked: 2026-08-21
Artifact identity does not establish licence or context truth. Those checks remain separate.
Variants
Gemma models grouped by workflow
Latest / flagship
Gemma 4
Google · Gemma
Best for: Developers evaluating Google-backed open-weight models for efficient local apps, hosted prototypes, and multimodal workflows where supported.
Local: Smaller Gemma variants are practical for local testing; larger variants need more VRAM or unified memory.
Gemma 3 4B IT
Google · Gemma
Best for: Efficient local prototypes, app workflows, and Gemma-family comparisons.
Local: Use the exact checkpoint and quantization that matches your hardware and latency target.
Gemma 4 31B
Google · Gemma
Best for: Maximum Gemma quality for local chat, reasoning, and long-context work on 24 GB cards.
Coding
Vision / multimodal
Local-friendly
Gemma 3 12B IT
Google · Gemma
Best for: Efficient local prototypes, app workflows, and Gemma-family comparisons.
Local: Use the exact checkpoint and quantization that matches your hardware and latency target.
Gemma 3 1B IT
Google · Gemma
Best for: Efficient local prototypes, app workflows, and Gemma-family comparisons.
Local: Use the exact checkpoint and quantization that matches your hardware and latency target.
Gemma 2 27B Instruct
Google · Gemma
Best for: Efficient local prototypes, app workflows, and Gemma-family comparisons.
Local: Use the exact checkpoint and quantization that matches your hardware and latency target.
Gemma 2 9B Instruct
Google · Gemma
Best for: Efficient local prototypes, app workflows, and Gemma-family comparisons.
Local: Use the exact checkpoint and quantization that matches your hardware and latency target.
Gemma 2 2B Instruct
Google · Gemma
Best for: Efficient local prototypes, app workflows, and Gemma-family comparisons.
Local: Use the exact checkpoint and quantization that matches your hardware and latency target.
Compare
All Gemma models in the directory
| Model | Type | Best for | Local runner notes | License | Detail |
|---|---|---|---|---|---|
| Gemma 4 | Chat | Developers evaluating Google-backed open-weight models for efficient local apps, hosted prototypes, and multimodal workflows where supported. | Smaller Gemma variants are practical for local testing; larger variants need more VRAM or unified memory. | Apache 2.0 | Open |
| Gemma 3 12B IT | Edge | Efficient local prototypes, app workflows, and Gemma-family comparisons. | Use the exact checkpoint and quantization that matches your hardware and latency target. | Gemma Terms of Use | Open |
| Gemma 3 4B IT | Edge | Efficient local prototypes, app workflows, and Gemma-family comparisons. | Use the exact checkpoint and quantization that matches your hardware and latency target. | Gemma Terms of Use | Open |
| Gemma 3 1B IT | Edge | Efficient local prototypes, app workflows, and Gemma-family comparisons. | Use the exact checkpoint and quantization that matches your hardware and latency target. | Gemma Terms of Use | Open |
| Gemma 2 27B Instruct | Chat | Efficient local prototypes, app workflows, and Gemma-family comparisons. | Use the exact checkpoint and quantization that matches your hardware and latency target. | Gemma Terms of Use | Open |
| Gemma 2 9B Instruct | Chat | Efficient local prototypes, app workflows, and Gemma-family comparisons. | Use the exact checkpoint and quantization that matches your hardware and latency target. | Gemma Terms of Use | Open |
| Gemma 2 2B Instruct | Edge | Efficient local prototypes, app workflows, and Gemma-family comparisons. | Use the exact checkpoint and quantization that matches your hardware and latency target. | Gemma Terms of Use | Open |
| CodeGemma 7B | Code | Coding assistant experiments and developer workflow prototypes. | Use the exact checkpoint and quantization that matches your hardware and latency target. | Gemma Terms of Use | Open |
| PaliGemma 2 | Vision | Vision-language app prototypes and multimodal evaluation. | Pick the resolution variant to match the task: 224px for natural images, 448px or 896px when the answer depends on small text such as documents and charts. | Gemma Terms of Use | Open |
| Gemma 4 31B | Chat | Maximum Gemma quality for local chat, reasoning, and long-context work on 24 GB cards. | Ollama Q4_K_M is ~18 GB (26B A4B) / ~20 GB (31B): full-GPU on 24 GB cards, split placement with reduced throughput on smaller GPUs. | Apache 2.0 | Open |