Gemma 4 31B

This model excels at multimodal reasoning across screenshots, documents, diagrams, and design assets. Ideal for visual agentic workflows, image-aware copilots, and teams migrating from closed multimodal APIs to an open model.