Introduction

What is Tera Router?

Tera Router is an LLM gateway you run yourself. Point your SDK, curl, or favorite coding agent at a single base URL, and the router forwards each request to the providers you have connected — complete with fallbacks, cooldowns, spending limits, and usage tracking.

• Three dialects, fully compatible — POST /v1/chat/completions (OpenAI), POST /v1/messages (Anthropic), and POST /v1/responses (OpenAI Responses). Existing clients need no changes.
• Routing you control — provider chains with priority, round-robin, or load-balanced strategies, plus automatic failover when an account fails or gets rate limited.
• Cost control — plans and budgets per key, per account, or global, with alerts and hard cutoffs.
• Guardrails — PII, prompt injection, topic, toxicity, and bias detection before requests leave.
• Dashboard — manage everything above from a UI: API keys, provider accounts, chains, guardrails, usage, down to the console log.

Three steps to your first request

1. Start the server — make run in apps/server, gateway ready at http://localhost:8080.
2. Create an API key — sign in to the dashboard at http://localhost:5173, open API Keys, create a key prefixed with sk_tr_.
3. Call it — use that key in the Authorization header:

curl http://localhost:8080/v1/chat/completions \
  -H "Authorization: Bearer sk_tr_..." \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-4o-mini",
    "messages": [
      { "role": "user", "content": "Hello, world!" }
    ]
  }'

Call it from code

Any SDK that speaks the OpenAI format only needs its base URL swapped:

import OpenAI from 'openai'

const client = new OpenAI({
  apiKey: 'sk_tr_...',
  baseURL: 'http://localhost:8080/v1',
})

const res = await client.chat.completions.create({
  model: 'anthropic/claude-sonnet-4',
  messages: [{ role: 'user', content: 'Hello, world!' }],
})

import Anthropic from '@anthropic-ai/sdk'

const client = new Anthropic({
  apiKey: 'sk_tr_...',
  baseURL: 'http://localhost:8080',
})

const res = await client.messages.create({
  model: 'openai/gpt-4o-mini',
  max_tokens: 1024,
  messages: [{ role: 'user', content: 'Hello, world!' }],
})

Note: with Tera Router, an Anthropic model can be called through the OpenAI SDK and vice versa — the router translates the dialects.

Next steps