gemma4:e4b

A quantized 4-bit execution of Google’s lightweight Gemma model optimized for local inference via Ollama.

The gemma4:e4b tag identifies a specific 4-bit quantized version of Google’s open-weights architecture designed for high-speed local deployment. By utilizing the e4b quantization method (a variant of the Q4 format), the model reduces memory overhead while maintaining high reasoning accuracy for text generation and code completion tasks. It runs efficiently on consumer hardware with 8GB of VRAM or less, making it a primary choice for developers building low-latency AI assistants or automated pipelines.

https://ollama.com/library/gemma