Skip to content

AI Control Plane Platform

Unified AI infrastructure for the enterprise. One API, 100+ models, 9 providers, intelligent routing, and complete cost visibility.

What is the AI Control Plane?

The AI Control Plane is a self-hosted platform that sits between your applications and AI providers (OpenAI, Anthropic, Google, xAI, DeepSeek, AWS Bedrock, Azure, Vertex AI, Ollama). Every request flows through a single OpenAI-compatible API endpoint, giving you:

  • Provider abstraction — switch models with a one-word change, no code rewrites
  • Intelligent routing — automatic failover, load balancing, and policy-based model selection
  • Cost management — real-time spend tracking, per-team budgets, and alerts
  • Enterprise controls — RBAC, rate limiting, audit logging, and Cedar policy engine
Guide Description
Quickstart Get running locally in under 5 minutes
API Integration Code examples in Python, TypeScript, Go, and curl
Model Routing How models are selected, fallback chains, and group aliases
Cost Management Budgets, alerts, FinOps reporting, and optimization tips
Admin Guide Page-by-page walkthrough of the Admin Console

Architecture Overview

Your App ──▶ AI Control Plane (LiteLLM) ──▶ OpenAI / Anthropic / Google / xAI / DeepSeek / Bedrock / Azure / Vertex / Ollama
                 │
                 ├── Agent Gateway (MCP + A2A protocols)
                 ├── Policy Router (Cedar rules)
                 ├── Semantic Cache (Redis + embeddings)
                 ├── Cost Predictor (per-request estimates)
                 ├── Budget Webhook (soft/hard limits)
                 ├── FinOps Reporter (CSV/JSON exports)
                 ├── Workflow Engine (LangGraph templates)
                 ├── A2A Runtime (Temporal agent workflows)
                 └── Admin API + UI (configuration & monitoring)

Default Services

Start the platform with docker compose --env-file config/.env up -d:

Service URL Purpose
LiteLLM localhost:4000 OpenAI-compatible LLM proxy
Admin UI localhost:5173 Web admin console
Landing Page localhost:9999 Interactive demo & playground
Docs Site localhost:8089 Developer documentation
Admin API localhost:8086 REST management API

Production URLs

Service URL
LLM API https://api.aicontrolplane.dev
Admin Console https://api.aicontrolplane.dev/admin
Developer Docs https://docs.aicontrolplane.dev

Model Groups

Request a group alias and the gateway picks the best available model:

Alias Models Best For
fast gpt-5-mini, claude-haiku-4.5, gemini-3-flash, grok-3-mini Low-latency responses
smart gpt-5, claude-sonnet-4.5, gemini-3-pro, grok-4 Balanced quality & speed
powerful gpt-5.2, claude-opus-4.5, o3-pro, grok-4-heavy Maximum capability
reasoning o3, o3-pro, deepseek-r1 Complex multi-step reasoning
coding claude-sonnet-4.5, deepseek-coder, codellama Code generation & review
cost-effective gpt-5-mini, claude-haiku-4.5, gemini-2.5-flash-lite, deepseek-v3 Budget-friendly 
# One API, any model
curl http://localhost:4000/v1/chat/completions \
  -H "Authorization: Bearer sk-litellm-master-key-dev" \
  -H "Content-Type: application/json" \
  -d '{"model": "smart", "messages": [{"role": "user", "content": "Hello!"}]}'