CrayEye

CrayEye is a multimodal multitool for working with visual AI models. It lets a user craft and share multimodal LLM vision prompts that are infused with real-world context pulled from a device's sensors and APIs, such as location and weather. CrayEye is free, open-source, and built entirely through AI-driven software development.

A user interprets their surroundings by pointing a smartphone camera at a scene while CrayEye uses AI to analyze what the camera sees. Prompts can be customized with context such as the device's location or current weather, pulled in through connected APIs, so each analysis reflects the user's actual environment rather than a generic description.

CrayEye includes a set of featured prompts that demonstrate what the multitool can do, and a demo reel showing examples of its visual analysis in action. Users can edit existing prompts or create new ones, then share them with friends, who can in turn add their own prompts to the collection. The project is open-source, with its source code and the story of how it was built both published alongside the tool.

Use cases

• Interpreting a physical environment through a smartphone camera
• Experimenting with visual multimodal models in new ways
• Browsing featured community prompts for inspiration