unsloth/gemma-3-12b-it-qat

The unsloth/gemma-3-12b-it-qat model is a 12 billion parameter instruction-tuned multimodal language model from Google DeepMind, part of the Gemma 3 family. It supports both text and image inputs with a 128K token context window and generates text outputs. This specific variant utilizes Quantization Aware Training (QAT) to maintain bfloat16 quality while significantly reducing memory requirements, making it suitable for resource-constrained environments. It excels in text generation, image understanding, and reasoning tasks across over 140 languages.