Raeto routing · for every app

Cut your AI bill.
Keep your keys.

Raeto picks the right model for every request your app makes. Simple questions go to a fast, cheap model; hard ones go to a strong one. You call the model with your own provider key, so nothing about your setup changes.

Get an API key See what you'd save

Up to 70% off simple requests · works on any chain · $0.0002 per decision

Your app asks RaetoSend the chat request you're about to make.

Raeto picks the modelBased on what the request needs: size, code, tools, images, reasoning.

fast: gpt-4o-minimid: gpt-4.1top: o3

You call it yourselfWith your own OpenAI, Anthropic or OpenRouter key. We never see the key or the reply.

How it works

Two lines. That's the integration.

Wrap the OpenAI client you already use. Simple requests go to a cheaper model from the same family, and everything else stays on the model you picked. If Raeto is ever slow or unreachable, the request goes to your usual model, so it can't break your app.

1
Get a keyOne click. No sign-up, no wallet.
2
Add the SDKOne file, no dependencies. Node, Bun, Deno or Python.
3
Wrap your clientThen use it exactly as before.

// curl -o raeto-route.mjs https://raeto.cc/sdk/raeto-route.mjs
import OpenAI from "openai";
import { withRaeto } from "./raeto-route.mjs";

const openai = withRaeto(new OpenAI(), { apiKey: process.env.RAETO_API_KEY });

// use openai exactly as before
const reply = await openai.chat.completions.create({ model: "gpt-4.1", messages });

# curl -o raeto_route.py https://raeto.cc/sdk/raeto_route.py
import os
from openai import OpenAI
from raeto_route import with_raeto

client = with_raeto(OpenAI(), api_key=os.environ["RAETO_API_KEY"])

# use client exactly as before
reply = client.chat.completions.create(model="gpt-4.1", messages=messages)

# Ask which model to use, then call it with your own provider key.
curl https://api.raeto.cc/v1/route \
  -H "Authorization: Bearer $RAETO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"messages":[{"role":"user","content":"Summarize this ticket"}],
       "models":{"fast":"gpt-4o-mini","mid":"gpt-4.1","top":"o3"}}'

# → {"model":"gpt-4o-mini","tier":"fast","reasons":["simple request"],"fallbacks":["gpt-4.1","o3"]}

01Any providerName your own model for each tier: OpenAI, Anthropic, Google, OpenRouter, or your own.

02Any chainIt's a plain HTTPS API. Robinhood Chain, Base, Solana, or no chain at all.

03Nothing to migrateKeep your SDK, your account and your billing with your provider. Add one call.

Savings calculator

What would you save?

Most apps send every request to one strong model. Plenty of those requests are simple and don't need it.

Requests per month Model you use today

Input tokens per request Output tokens per request

Requests that are simple: 60%Those go to a fast model. The rest stay on your current model.

Estimated savings per month

Today$0

With Raeto routing$0

Routing fees included$0

Cut0%

An estimate using list prices per million tokens; check your provider's current prices. Your real mix will differ. Simple requests are priced at GPT-4o mini.

Built for

Anything that calls an AI model.

Telegram & Discord botsMost messages are short. Stop paying top-model prices for "gm".

LaunchpadsToken descriptions, summaries and moderation at a fraction of the cost.

DEXes & perpsAI trade explainers and assistants that scale with volume, not with your bill.

AI apps & agentsRoute every step of an agent to the cheapest model that can do it.

Your setup stays yours

We pick the model. That's it.

Never sees your keysYou call the model yourself. Raeto never touches your provider account.

Never sees repliesThe request you send is used only to pick a model, and isn't stored.

Can't break your appIf Raeto is ever slow or down, requests go to your usual model. Your key can only pick models, so it can't spend anything.

Pricing

Pay per decision.

Routing only

$0.0002 per routed request

$0.20 per 1,000 requests, whatever their size
Your provider bills you for the model as usual
No monthly fee, no minimum, no $RAETO needed

Prefer one bill?

Raeto API

Skip your own provider: call 400+ models through Raeto
Pay from your Raeto balance at the model's price plus a small fee
See the developer docs

Top up your Raeto balance with USDG, a card, or rCREDIT. Revenue from routing is Raeto revenue, which staking will share with holders once it launches.

Start in a minute

Route your first request today.

Get a key in one click, add two lines, and your next request is routed. No sign-up, no wallet.

Get an API key Read the docs