This roadmap reflects intent, not promises.
- Local image generation backend
- Web UI with live progress
- Multiple built-in models
- NSFW filtering
- Gallery & asset system
- Improved error handling
- Better WebSocket resilience
- Unified settings system
- English-only UI consistency
- Improved documentation
- React frontend implementation
- Hardware monitoring (CPU, RAM, GPU)
- Multiple upscaler methods
- NVIDIA CUDA — Full support
- AMD ROCm — Experimental support
- Intel Arc XPU — Experimental support (oneAPI)
- Apple Silicon MPS — Experimental support
- CPU fallback — Always available
- Runtime device switching
- DRAM Extension for low-VRAM GPUs
- Admin Panel — Protected server management UI
- Server Modes — api, html, react, full
- Image Variations — img2img support
- Autosave Settings — No manual save needed
- CTRL+R Restart — Quick server restart
- Enhanced Monitoring — GPU temp, power, utilization
- Multi-user support (optional)
- Token / quota system (optional)
- Headless API mode
- External UI integrations
- Batch generation queue
- Plugin system
- Community model packs
- Custom UI themes
- Reference deployment examples
- Docker containerization
This roadmap will evolve with contributors.