Inference routing core
ORTHO AI Gateway
Problem
An explicit inference request path across the ORTHO stack.
Approach
M/MUMPS routines implementing request state, routing, provider calls, streaming events, IPC, and usage records.
Verification and limits
The public README describes the current M/MUMPS core. Website chat is a separate, restricted local-model gateway; it does not expose administrative tools.
This project summary reviews public documentation. It does not substitute for the project's build, tests, proof audit, or independent evaluation.
License
BSL / AGPL / MPL — see terms. The actual project terms govern use.
Read project terms ↗