context-labs/google-gemma-3-27b-it

Gemma 3, developed by Google DeepMind, is a family of lightweight, open multimodal models built from the same research as Gemini. The 27 billion parameter instruction-tuned variant handles text and image input, generating text output, and features a 128K context window. Optimized for a variety of text generation and image understanding tasks, including question answering, summarization, and reasoning, it supports over 140 languages. Its relatively small size allows for deployment in resource-limited environments like laptops and desktops.