Are requests sent directly to providers or through aggregator chains?

Requests can be sent on direct provider paths, and TheRouter exposes that routing evidence instead of hiding it behind an opaque chain.

How can teams verify which provider actually served a request?

Each response can include routing headers, and teams can inspect the request trace in the dashboard to see which provider path was selected.

Can we tell when failover happened?

Yes. Transparent routing shows whether a fallback fired, how long each attempt took, and why a different path was selected.

Why does transparent routing matter for enterprise teams?

Transparent routing reduces hidden vendor risk and makes reliability, cost review, incident analysis, and compliance conversations easier because teams can inspect real routing evidence instead of relying on claims.

Transparent Routing

Know Exactly Where Your Request Goes

Every response tells you which provider handled your request. Direct connections only — no aggregators, no middlemen, no black boxes.

Get Started See How It Works

The Industry Has a Trust Problem

Most API gateways route through chains of aggregators. You never know who actually served your request.

Typical API Proxy

Your App

Proxyunknown

Aggregatorcached?

Sub-aggregatorwhich model?

???

Provider

TheRouter

Your App

TheRouterverified

Provider (Direct API)x-router headers

Three Pillars of Transparency

aggregators

Direct Connections Only

We connect to Anthropic, OpenAI, Google, AWS Bedrock directly. No middlemen reselling API access.

Every provider is a first-party API connection
No resold, cached, or proxied responses
Your data touches only you and the model vendor

x-router-* headers

Verifiable Response Headers

Every API response carries headers that tell you exactly what happened: which model, which provider, whether a fallback fired, and how long the upstream took.

x-router-model: the model that served your request
x-router-provider: the verified direct provider
x-router-fallback: whether a fallback was used

100%

requests traced

Visual Routing Trace

Open any request in your dashboard and see the full routing graph: which providers were tried, which was selected, timing for each attempt, and why.

Flow graph: model → candidates → selected provider
Per-attempt timing and failure reasons
Full audit trail stored for every request

See It in Action

Every response carries routing proof. Here’s what a real API call looks like:

terminal

$ curl -s -D- https://api.therouter.ai/v1/chat/completions \
  -H "Authorization: Bearer sk-your-key" \
  -H "Content-Type: application/json" \
  -d '{"model":"anthropic/claude-sonnet-4.6","messages":[{"role":"user","content":"Hi"}]}'

HTTP/2 200
x-router-request-id: req_a1b2c3d4e5
x-router-model: anthropic/claude-sonnet-4.6
x-router-provider: Anthropic
x-router-fallback: false
x-router-latency-ms: 847

{"id":"chatcmpl-...","model":"anthropic/claude-sonnet-4.6",...}

Transparent Routing FAQ

Verify provider path, failover, and routing evidence on every request.

Want more routing detail? Read the model fallbacks guide and provider selection guide.

Start building in 3 lines of code

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.therouter.ai/v1",
  apiKey: "sk-your-key",
});

const response = await client.chat.completions.create({
  model: "anthropic/claude-sonnet-4.6",
  messages: [{ role: "user", content: "Hello!" }],
});

Get API Key OpenAI-Compatible Quickstart