SharedLLM
One CLI for every coding agent
Any model, one command
Run Claude Code through Anthropic, OpenAI, Ollama, Kimi and more — switch with a flag, no config rewrites.
Auto smart routing
Auto resolves each request to the first model with an available key — your own first, then the shared pool.
Bring your own keys
Use your existing provider keys. Encrypted at rest, never shared without your say-so.
Full observability
Every request metered with per-developer spend caps and a unified usage dashboard.
SharedLLM · Transparent AI gateway