Self-hosted LLM gateway

One base URL for every model.

Point your SDK, curl, or favorite coding agent at a single endpoint. The router takes care of providers, fallbacks, spend, and guardrails.

Quick start Explore the Gateway API

API dialects: 3
providers out of the box: 20+
fallback targets: ∞

Terminal

$ curl http://localhost:8080/v1/chat/completions \
  -H "Authorization: Bearer sk_tr_..." \
  -H "Content-Type: application/json" \
  -d '{
      "model": "anthropic/claude-sonnet-4",
      "messages": [{ "role": "user", "content": "Hi!" }],
  }'

X-TeraRouter-Provider:anthropic

Explore the documentation

From zero to your first request, then everything in between.

Getting started

Run the server and dashboard locally within minutes.

Gateway API

Three dialects at once — OpenAI, Anthropic, and Responses API.

Models & Routing

Alias pools, chains with fallbacks, and reasoning suffixes.

Guardrails

PII, prompt injection, topics, toxicity, and bias — layered.

API Reference

Every dashboard REST endpoint for automation and integrations.

Configuration

Environment variables, make targets, and built-in limits.