gemma-3-12b-it

Model overview

gemma-3-12b-it is published by RedHatAI as an image-text-to-text model with Gemma3ForConditionalGeneration architecture. Safetensors metadata reports 12,187,325,040 parameters, or about 12.19B, and hub tags record a google/gemma-3-12b-pt base. The model card describes the Gemma 3 family as multimodal text-and-image input models with 128k context and broad multilingual coverage.

Recorded capabilities

Use cases in the source record

• Question answering, summarization, and reasoning over text and image inputs, which the model card names as suited tasks for Gemma 3.
• Vision-language experiments across industries and domains, framed by the model card as contextual possible uses rather than an exhaustive list.

Limitations and unknowns

• No evaluation results were extracted from this record.
• Provider state is historical snapshot data and should be refreshed before being presented as current.
• No VRAM, quantization, pricing, or current deployment figures were extracted.
• The model card notes limitations including training-data influence on capabilities and possible incorrect or outdated factual statements.

Source and provenance

Source: RedHatAI/gemma-3-12b-it

Captured: Unknown. Processed: 2026-09-07T19:35:51.336566+00:00.

Gemma 3 model card Model Page : Gemma Resources and Technical Documentation : Gemma 3 Technical Report Responsible Generative AI Toolkit Gemma on Kaggle Gemma on Vertex Model Garden Terms of Use : Terms Authors : Google DeepMind Model Information Summary description and brief definition of inputs and outputs. Description Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models. Gemma 3 models are multimodal, handling text and image input and generating text output, with open weights for both pre-trained variants and instruction-tuned vari…