When to use Prompt to Video - Velo Documentation

Prompt to Video is one of the ways to make a video in Velo. It shares everything the other paths use — workspaces, voice and face clones, libraries, sharing, and analytics. The only difference is what you start from.

Start from an asset

Use one of the recording, upload, or link paths when you already have something to work from — a live screen recording, an uploaded video or PDF, or a link to a website or app. Velo turns that existing asset into a narrated video. See Creating a video for the available paths.

Start from a prompt

Use Prompt to Video when you have no asset yet — just an idea of what needs explaining. You describe the video in a sentence or two, and the agent researches, scripts, voices, and renders it for you, pulling in context from your connected knowledge sources along the way.

Choosing between them

Both produce a video in Velo. They differ only in where the video starts and how it is billed.

	Start from an asset	Start from a prompt
Starting point	A recording, upload, or link	A text prompt
Best for	Explaining something that already exists (a demo, a document, a product)	Explaining something you can describe but haven’t recorded yet
Knowledge sources	The asset you provide	Knowledge bases, MCP servers, and connected applications
Editing	Scene-by-scene editor, with timeline and effects	Prompt-based re-editing only — no scene editor
Billing	Velo minutes (shared pool)	Prompt to Video credits (separate pool)

Sharing, analytics and VeloTwin work the same either way. Editing does not: only asset-based videos open in the scene editor. If you’re not sure which to use, ask: do I have something to record or upload, or do I only have something to describe? That answers it.

⌘I