| Local model runtime |
Current: configured llama-server executable for local GGUF models |
Future platform foundation |
| Admin and user model behavior |
Current: admins load/stop/configure models; users chat with the loaded main model |
Future platform boundary |
| Streaming chat workspace |
Current: streaming, stop, regenerate, prompt edit, latest-turn variants, Markdown, code, math, and reasoning display |
Expected core workflow in future platform work |
| File-aware conversations |
Current: supported text/code attachments with server-side limits, chunking, and context budgeting |
Expected core workflow in future platform work |
| Runtime Visibility |
Current: live llama-server Logs, runtime status, GPU Monitor data through local NVIDIA/AMD tools where available, Analytics, and tensor-split launch support where the runtime allows it |
Broader multi-system visibility is a future platform direction |
| Benchmarks |
Current: admin-only CE prompt set, eligibility controls, live progress, best-run tracking, and details |
Expanded benchmark systems are a future expansion area |
| Installation & Settings |
Current: first-run installer, runtime path settings, initial admin creation, users, and local configuration |
Larger deployment workflows are a future platform direction |
| Image understanding / generation |
Planned CE roadmap; not current v1.0 functionality |
High-level platform direction |
| API support |
Planned CE roadmap; not current v1.0 functionality |
Future automation and orchestration direction |
| Multi-node / distributed inference |
Not current CE functionality |
Future Pro/platform concept |
| Cluster-aware infrastructure |
Not current CE functionality |
Future Pro/platform concept |