XRAI (eXtended Reality AI) is an open universal file format designed for real-time interactive spatial media that seamlessly bridges current technologies with future AI-driven XR experiences. It enables dynamic, collaborative 3D/4D content delivery across web, mobile, AR, VR, and emerging wearable devices.
This repository contains the technical specification and example implementations for the XRAI format, a next-generation standard for AI+XR spatial media that is:
- Fast & Lightweight: Optimized for real-time streaming and rendering
- AI-Driven: Native support for neural networks, procedural generation, and context-aware adaptation
- Cross-Platform: Compatible with web, mobile, AR, VR, and future devices
- Future-Ready: Designed for emerging technologies like neural interfaces and holographic displays
- Open Standard: Community-driven development with transparent governance
spec/: Technical specification documentscore-spec.md: Core format specificationvfx-spec.md: VFX system specificationai-components.md: AI integration specification
examples/: Example implementationsweb/: Web viewer implementationunity/: Unity integration
schemas/: JSON schemas for format validation
- Hybrid representation system combining explicit geometry with neural fields
- Advanced VFX capabilities for dynamic particle systems and shaders
- AI-driven procedural content that adapts to context
- Optimized for current and future hardware capabilities
- Collaborative features for shared spatial experiences
See the Core Specification for a detailed overview of the format structure and capabilities.
For implementation examples, check the Examples directory.
This specification is open source and available under the MIT License.