Question 1

Can LiteLLM run Claude Code, Codex, or Hermes as agents?

Accepted Answer

LiteLLM routes model completion requests — it does not boot agent harnesses or manage cloud compute. Claude Code, Codex, and Hermes are execution environments that interpret model output and take file-system and tool actions inside an isolated computer. You could use LiteLLM to route the model calls those harnesses make internally, but you would still need to manage the harness lifecycle, compute, and state yourself. AgentSky ships all of that behind one call.

Question 2

LiteLLM is open-source and free to self-host — why use AgentSky instead?

Accepted Answer

LiteLLM's free tier routes model requests; you still supply the compute, the agent loop, the harness, persistent state, and crash recovery. AgentSky is not a model router — it is a managed agent runtime that ships the harness, provisions the computer, and handles recovery. They solve different problems: LiteLLM is the routing layer below an agent; AgentSky is the agent execution layer above the model. Teams that self-host LiteLLM for model routing can also call AgentSky for the agent sessions.

Question 3

Does LiteLLM support BYO model subscriptions the way AgentSky does?

Accepted Answer

LiteLLM routes at the API level — it calls provider APIs and bills per token at those rates. AgentSky supports BYO Claude and ChatGPT subscriptions that reduce model cost to $0 for those harnesses by using subscription capacity rather than API tokens. That BYO subscription model is specific to AgentSky's relationship with the harness providers.

Question 4

I use LiteLLM for cost tracking — can I use AgentSky alongside it?

Accepted Answer

Yes. AgentSky tracks usage per session and per task through its own billing; LiteLLM tracks per-completion costs from model providers. If your stack uses LiteLLM to route model calls for non-agent workloads, adding AgentSky for agent sessions means maintaining two cost centers, but they are complementary and independent. AgentSky does not currently integrate with LiteLLM's budget-enforcement layer.

	Cloud agent platformAgentSky	LLM proxy and routerLiteLLM
What you get	A running cloud agent — harness, model, persistent compute, and event stream — behind one API call	A unified LLM completion interface with routing, fallback, cost tracking, and caching across 100+ providers
Primary layer	Agent runtime: provisions compute, boots harness, manages workspace, streams events	LLM proxy: translates and routes model requests; does not run agents or manage compute
Open-source / self-hostable	Managed cloud service; no self-host option	Core library and proxy are open-source (MIT); self-hostable on any infrastructure; Enterprise tier is managed
Agent harness	8 maintained harnesses shipped and operated by AgentSky	Not included — LiteLLM routes model calls; you build the agent loop yourself
Persistent compute	Isolated cloud computer per session; crash-resistant, resumes mid-task automatically	Not provided — stateless model routing only; compute is your infrastructure responsibility
Cost tracking	Usage tracked per session; BYO subscription reduces model cost to $0	Per-request cost tracking and budget limits built in; alerts configurable per team or key
Pricing model	Pay per task at list pricing; no card required to sign up	Open-source version is free (self-hosted); LiteLLM Proxy+ Enterprise pricing on request
Best for	Running complete agentic coding tasks immediately without building agent infrastructure	Teams that need a unified LLM interface, provider-level routing, and cost governance across many providers

AgentSky vs LiteLLM

How they compare

LiteLLM is the right choice for these workloads

Common questions

The agent and its computer. One API.