Roadie · by Paroaria
One integration for multiple model providers — chat, embeddings, and images — with the authentication, limits, and observability an app actually needs. Bring your own provider keys.
// Keep the OpenAI SDK you already use — // change two lines and route through Roadie. import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://gateway.roadie.paroaria.ai/openai/v1", apiKey: process.env.ROADIE_KEY, // rd_sk_… }); const reply = await client.chat.completions.create({ model: "openai/gpt-4o", messages: [{ role: "user", content: "Hello" }], });
What Roadie does
Roadie sits between your app and the AI providers, so you ship features instead of infrastructure. Everything below works today.
Reach OpenAI and Anthropic through one API, with automatic fallback when a provider stumbles.
Chat · Embeddings · ImagesYour provider keys stay encrypted and are only ever decrypted at the gateway — never exposed to your app or ours.
Envelope-encryptedShip on-device image generation with RoadieKit and Apple App Attest — no secret key ever embedded in your app.
RoadieKit · SwiftSet rate limits, monthly quotas, and spend caps — per project or per end user — so cost and abuse stay in bounds.
Per-user controlsSee requests, tokens, and estimated provider cost, plus per-request logs — with privacy modes that redact what you don't want stored.
ObservabilityAttribute every request to a user, entitle free and pro tiers, and block abuse instantly — the backend for an app, not just an API key.
Free → ProStart here
Read how it works, create an account, or jump straight into the console to wire up your first project.