The fleet
PRD-006UnitLive

Cortex

LLM gateway & routing · AI Infrastructure

One endpoint for every model — route by cost, latency or quality, with caching and fallback built in.

40%
lower token cost
release_position
Coming soon
Early access
Beta
Live

Cortex is currently live on the Diathic fleet — where this unit sits on the road to general availability.

datasheet

Spec dump for Cortex — the numbers before the pitch.

designation
Cortex
unit id
PRD-006
class
AI Infrastructure
status
Live
interface
REST · SDK · OpenAI-compatible
deployment
Managed · self-host
pricing
Per token routed
headline
40% · lower token cost
subsystems
MOD_01active

One endpoint for every provider — route each request by cost, latency or quality.

MOD_02active

Response caching and automatic fallback the moment a provider degrades.

MOD_03active

Per-team budgets, rate limits and spend alerts built in.

MOD_04active

Full request logs and analytics, wired to Sentinel out of the box.

Put Cortex to work

Tell us about your stack and we'll set up a walkthrough — or get you into the beta.

Start a project