Skip to main content

Logos Documentation

Logos

LLM engineering made easy

OpenAI-compatible inference with logging, billing, policies, and self-hosted workers — documented for users, operators, and contributors.

Choose a guide

Four paths through the same platform — pick the one that matches what you are doing today.

How Logos fits together

Clients hit one domain. The web service terminates the public inference gateway; local and mixed traffic is proxied to the orchestrator; workers dial out over WebSocket.

Logos top-level system design — clients, core node, workers, and cloud providers

Top-level design, deployment, data model, and request flow live in System Design. Terminology is in the architecture reference.