Gemma 3n E2B
by Google · gemma-3 family
2B
parameters
text-generation reasoning multilingual vision
Gemma 3n E2B is Google's ultra-compact 2B-parameter multimodal model from the Gemma 3n family, optimized for edge and on-device inference. It handles both text and vision inputs while requiring minimal memory and compute resources. As the smallest model in the Gemma 3n lineup, the E2B variant is ideal for resource-constrained environments where a lightweight model with multimodal capabilities is needed, such as mobile devices and IoT applications.
Quick Start with Ollama
ollama run e2b-q4_K_M | Creator | |
| Parameters | 2B |
| Architecture | transformer-decoder |
| Context | 32K tokens |
| Released | Sep 10, 2025 |
| License | Gemma Terms of Use |
| Ollama | gemma3n:e2b |
Quantization Options
| Format | File Size | VRAM Required | Quality | Ollama Tag |
|---|---|---|---|---|
| Q4_K_M rec | 1.3 GB | 3.3 GB | | e2b-q4_K_M |
| Q8_0 | 2.2 GB | 4.2 GB | | e2b-q8_0 |
| F16 | 4 GB | 6.5 GB | | e2b-fp16 |
Compatible Hardware
Q4_K_M requires 3.3 GB VRAM