CrayEye

CrayEye is an open-source multimodal tool that lets people craft and share vision prompts for large language models, augmented with real-world context pulled from a phone's sensors and APIs, such as location and weather. Pointed at the camera, it feeds an image plus a chosen prompt to a multimodal LLM and returns an AI-generated interpretation of what the camera sees, shaped by whatever context the prompt requests. Users can browse and reuse a library of community-submitted 'Featured Prompts' or write their own, which makes CrayEye function partly as a prompt-sharing platform for visual AI rather than just a single-purpose camera app. It's built for developers and AI enthusiasts who want to experiment with combining vision models and live sensor data, and its source code is published on GitHub for anyone who wants to inspect or extend it. The app is available on iOS and Android and is free to use. Because CrayEye is a thin interface over third-party multimodal LLMs rather than a model it trains itself, output quality and accuracy depend on whichever underlying model is configured -- a limitation inherent to any prompt-layer tool built on top of external AI providers.