Isaac 0.1, the first model in the Perceptron family, represents a significant advancement in AI's ability to understand and interact with the physical world by providing capabilities in grounding, spatial reasoning, and in-context visual learning. Despite having only 2 billion parameters, Isaac 0.1 surpasses much larger models in perceptive benchmarks, making it suitable for real-time and edge deployments. It excels in applications such as visual question answering, pointing and localization accuracy, and learning new visual concepts from minimal examples. Designed for real-world applications, the model can identify missing PPE, detect defects in manufacturing, and support security use cases by spotting anomalies. Additionally, it can be used in interactive experiences like AR-style cooking assistants and real-time repair guidance. Developers can try Isaac 0.1 through fal or Perceptron's demo page, allowing them to experiment with images and queries to explore its capabilities.