Cortega AI Gateway governs the AI traffic behind products, internal services, agents, model gateways, and MCP tool calls, giving platform, security, compliance, and business teams a single control point for production use. It's free to run, with no time limit, for a real technical trial.
Every design partner and integration partner who has installed Cortega says the same thing: it's straightforward, painless, and fast, typically under 15 minutes from pulling the images to a running gateway.
Cortega ships as prebuilt Docker images on GHCR. No source build, no compiling, no dependency wrangling.
A single runtime .env file covers the setup. Provider API keys are optional and can be added later.
Sign in at the default admin login, change the password, and you have a live, governed AI gateway.
Cortega ships as a set of Docker images run with Docker Compose. The same images and compose stack work whether you're installing on a host you already manage, or using Cortega's scripted AWS or Terraform paths.
| Requirement | Details |
|---|---|
| Container runtime | Docker Engine + Docker Compose v2 on the host |
| Architecture | x86_64 or ARM64 (multi-arch images; Cortega's own scripted AWS install defaults to an ARM Graviton instance) |
| Recommended sizing to start | 2+ vCPU, 4+ GB RAM for a single-node trial; add more as gateway traffic grows |
| Network | Outbound reachability to ghcr.io to pull container images |
| Provider keys (optional) | API keys for whichever LLM providers you want to route to, added after install |
Every path runs the same images and the same Docker Compose stack. They differ only in how the infrastructure is created and how the runtime configuration is filled in.
Already have a Docker host, on-prem or in any cloud? Fill in a runtime configuration file by hand and run the compose files directly. No provisioning script involved.
Cortega provisions the EC2 instance and deploys onto it, for either a single-node or multi-node install.
Terraform provisions the infrastructure on AWS, Azure, or GCP, and a companion deployer installs the same stack.
Releases, install packages, and version notes live on GitHub.
Download softwareTell us about your environment and what you're trying to govern. We'll help you pick the right install path and sizing.