Bypassing MoA

Run only the raw transformer layer for lower latency when you don't need task-specific accuracy.

Turning it on

Skip MoA for a request with the x-interfaze-bypass-moa header:

x-interfaze-bypass-moa: true

If you use the Interfaze SDK, set bypassMoA: true when creating the client instead and the header is added for you. With any other SDK, set the header as a default header on the client so every request carries it.

import { Interfaze } from "interfaze";

const interfaze = new Interfaze({
  apiKey: process.env.INTERFAZE_API_KEY,
  bypassMoA: true, // run the raw model layer only
});

What you lose

With MoA bypassed, the model no longer runs the specialized tasks behind the scenes, so responses come back without precontext and the accuracy gains on OCR, web search, scraping, transcription, and other task-specific work go away.

If you want the raw output of one specialized task without the full model, run tasks is the better tool. It is both faster and cheaper than a full request, and you keep the specialized accuracy.