Skip to main content

How it works

You left a brief in a chat and got a video. Here is what actually happened, because knowing the shape makes everything else click.

The four stages

1. Plan

Your agent reads the brief and decides the structure of the piece: how many shots, what each shows, what the copy says, how long it holds. The plan is shaped by the agent’s flow — its defaults for medium, pacing, voice, and captions (see Flows).

2. Design

Each shot becomes a real, laid-out composition — headline, imagery, logo, spacing — in your brand’s fonts and colors automatically, because those live in the workspace as your brand DNA (see Brands). You never pick a font per piece; the brand decides. The result at this stage is a live composition you can preview, scrub through, and edit conversationally. Nothing is final until you render.

3. Render

Rendering turns the composition into a finished file — a video, an image, a document. This is real work done server-side, which is why it takes minutes and spends credits. Previews are cheap; renders are the investment.

4. Publish

A rendered piece is private until you publish it. Publishing puts it on your feed — and from the feed onto your wrapper site and connected channels. See Publishing.

Why this shape matters

The split between stages is the whole pricing and iteration model: Iterate above the line; render when it’s right; publish what you’re proud of.

Where agents fit

An agent is a configured content-maker: a brand identity, a flow, and a memory of its past work, scoped to one job — “our launch teasers”, “the weekly digest”. You will end up with a few. See Agents.