> ## Documentation Index
> Fetch the complete documentation index at: https://docs.orpheus.run/llms.txt
> Use this file to discover all available pages before exploring further.
# Overview
> Cognitive runtime for autonomous AI agents
**Orpheus is a specialized cognitive runtime engineered to provide long-term persistence, coordination, and governance for autonomous AI agents.**
Unlike traditional serverless functions or container orchestrators, it utilizes a three-layer architecture, coordination through workers, execution via lightweight **runc containers**, and persistence through dedicated workspaces.
The system optimizes for time and continuity rather than just latency, using **queue-depth autoscaling** to manage I/O-bound workloads that typically idle while waiting for LLM responses.
And the ServiceManager and Registry enable seamless integration with local inference engines like Ollama and vLLM, providing a platform-agnostic environment for complex reasoning systems.
With an actor model to eliminate race conditions and an ExecLog system for crash recovery and auditability.
## When to Use Orpheus
* AI agents that run for minutes (not milliseconds)
* Conversational agents that need context
* Agents that create or modify files
* Workflows that shouldn't lose state on restart
Deploy in 2 minutes
Step-by-step walkthrough