One OpenAI-compatible endpoint for every model you use.
One endpoint in front of every model.
Routing decisions recorded for every request.
Provider keys encrypted, scoped per project.
Deployed entirely on infrastructure you control.
Route requests across OpenAI, Anthropic, Google Gemini, and local models through a single OpenAI-compatible API — with per-project routing policy, bring-your-own-key, and budgets, self-hosted end to end.
Point any OpenAI SDK at one GaussMeridian endpoint. OpenAI, Anthropic, Google Gemini, and local runtimes like Ollama or vLLM all sit behind the same OpenAI-compatible API.
Set per-project policy over cost, quality floor, and model allowlists. Every routing decision is recorded, so you can see exactly why a request went where it went.
A web console manages API keys, project scoping, and monthly spend limits. Bring your own provider keys — encrypted at rest, managed per project.
Docker Compose brings up the gateway, SurrealDB, and Redis on infrastructure you control. Provider keys stay encrypted at rest, and routing, usage, and audit data stay in your own database — not a third-party SaaS.
Share a complex workflow. We’ll explore where a focused Digital Twin could improve visibility, test changes, and support better decisions.
Explore use cases