MODRO
Model Router

Hosted Routing Layer For OpenAI Traffic

Point your app at our hosted /v1. Keep your own OpenAI key. Cut waste automatically.

MODRO sits in front of your requests, routes simple work to cheaper models, preserves a high-quality path for harder prompts, and gives each tenant a private dashboard plus tenant-scoped model settings.

Quick Start How It Works See Demo

You bring your own OpenAI API key. Billing stays with your OpenAI account. We do not require new auth or account creation for the client flow.

One Small Integration

Full Setup Guide 
import os
import requests

BASE_URL = "https://modro.ai/v1"

response = requests.post(
    f"{BASE_URL}/chat/completions",
    headers={"Authorization": f"Bearer {os.environ['OPENAI_API_KEY']}"},
    json={
        "model": "balanced",
        "messages": [{"role": "user", "content": "Summarize the latest deploy notes."}]
    },
    timeout=30,
)

print(response.json()["choices"][0]["message"]["content"])
print(response.headers.get("X-Client-Dashboard-Url"))

From First Visit To Live Traffic

The user journey is intentionally short. Understand the product, paste one snippet, send traffic to hosted /v1, choose a preset, then open the signed private dashboard URL returned by the API.

1. Bring Your OpenAI Key

You keep your own OpenAI API key. This product does not replace your billing relationship.

2. Point To Hosted /v1

Your app sends requests to our hosted OpenAI-compatible endpoint. No full auth system or account migration required.

3. Open Your Private Dashboard

Each tenant gets a signed dashboard link in the response header. Use it to review savings and adjust model policy safely.

Presets And Model Roles

Users should not need to learn routing internals to get value. Presets stay simple, and the tier roles explain what each model is for.

Presets

  • Cheapest: push as much work as possible into the lowest-cost models.
  • Balanced: default for most teams, with strong quality and controlled cost.
  • Highest Quality: heavier fast and deep tiers for teams optimizing output quality first.

Tier Roles

  • Utility: trivial or low-value work where speed and cost matter most.
  • Fast: the default working tier for most traffic.
  • Deep: the expensive tier reserved for prompts that need more reasoning or coverage.
Utility Starting Point GPT-5 nano

$0.05 input / $0.40 output per 1M tokens. Best for the cheapest preset and low-stakes utility tasks.

Balanced Fast Tier GPT-5 mini

$0.25 input / $2.00 output per 1M tokens. Strong default fast tier for real traffic.

Highest Quality Deep Tier GPT-5.4

$2.50 input / $15.00 output per 1M tokens. Use when quality matters more than speed or price.

Why Teams Use It

Without a routing layer, expensive models end up handling a lot of routine traffic. That drives up cost and latency quietly. With tenant-scoped model policy and a private dashboard, the product stays easy to adopt and easy to trust.

Fast To Adopt

No full account system. No prompt rewrites. Point traffic to hosted /v1 and start.

Private By Default

Signed dashboard links preserve tenant isolation without exposing admin surfaces to end users.

Clear Next Step

Users can immediately open settings, choose a preset, and keep sending live traffic.

Start With The Quick Start

It explains the hosted endpoint, your own OpenAI key, the signed dashboard flow, the preset choices, and copy-paste Python and JavaScript snippets.

Open Quick Start Read Docs