AI infrastructure for products

One dependable layer between your product and AI.

Macrofix gives your application a clean, secure API for accessing the models you choose—without exposing upstream providers to every customer integration.

Customer API keys are securely hashed. Provider credentials stay on your server.

Request routed
POST /v1/chat/completions

{

"model": "macrofix/smart",

"messages": [

{ "role": "user", "content": "Help me build." }

]

}

One API

Your public surface

Model routing

Your decisions

Usage data

Your visibility

How it works

A clear path from your app to the right model.

01

Create a customer key

Issue scoped keys from your customer dashboard. The secret is revealed once; only its hash is retained.

02

Call one endpoint

Your product uses a consistent chat-completions API and your own public model aliases.

03

Operate with context

Track request volume, token use, model selection, status, latency, and upstream cost metadata.

Designed for control

Give customers simplicity without giving away your infrastructure.

Your team decides which models to expose, which provider handles them, and how the product presents them. Customers receive a single Macrofix key—not a collection of vendor accounts.

Provider isolation

Upstream credentials stay in server configuration.

Public aliases

Keep model IDs stable as your routing evolves.

Rate limits

Apply a clear request budget to every customer key.

Separate admin access

Customer and administrator sessions use different auth boundaries.

For developers

Build once. Route with intention.

Start with a customer workspace and make your first request through Macrofix’s OpenAI-compatible API surface.