> ## Documentation Index
> Fetch the complete documentation index at: https://docs.visualradioassist.live/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> Use the names from the glossary at https://docs.visualradioassist.live/glossary: Audio Manager (two words), command (never commando), on air and off air (never live) for programs and studios, and server apps for the VRA apps installed in the studio.
> Say which app an answer applies to, like VRA Cloud, the Core or the Output Player, and the version when the page names one.
> Only give settings, values and steps that are in these docs. Never invent one. If the docs don't cover it, say so.
> When the docs don't answer the question, send the user to Visual Radio Assist support at https://cloud.visualradioassist.live/support.

# Generate a graphic

> Let VRA build an editable first version of a graphic in your station branding, or rebuild an image of a graphic as editable layers. Each costs one AI credit.

The generator builds a first version of a graphic for you, as editable layers in the [Visual Designer](https://docs.visualradioassist.live/visual-designer.md). Pick a type and a style and VRA designs it in your [station branding](https://docs.visualradioassist.live/visual-designer/branding.md), or upload an image of a graphic and VRA rebuilds it. Each generated graphic costs 1 AI credit.

Go to **Visuals → Overview**, click **Add new** and choose **From generator**. **Add new** shows once the station has at least one visual.

## Generate a graphic from your branding

**1. Choose From your branding**

The generator needs your station branding. Without it, the page asks you to
set it up first.

**2. Pick a graphic type**

Choose **Lower third**, **Now playing**, **Ticker**, **Fullscreen**,
**Station bug** or **Clock**.

**3. Pick a style**

Choose **Bold**, **Modern** or **Classic**, and fill in the options of the
type, like the **Subject** and **Side** of a lower third.

**4. Generate**

Click the generate button, like **Generate lower third**. When it is done, the
Visual Designer opens on the new visual.

![](https://mintcdn.com/visualradioassist/cfPiQE-fjZkzRKZ8/images/visual-designer/generate-a-graphic/1.png?fit=max\&auto=format\&n=cfPiQE-fjZkzRKZ8\&q=85\&s=89535a2caad8a4bbd143e77ed4c5b00d)

## Generate a graphic from an image

Choose **From image** and upload an image of a graphic, like a screenshot or a design export. The image can be JPG, PNG, WEBP or SVG, at most 10 MB. Pick the **Aspect ratio**, **16:9 landscape** or **9:16 portrait**, and click **Generate visual**.

VRA detects the text, shapes and logos and creates editable layers for them, grouped into containers like a title tab or a news ticker.

## Review a generated graphic

A generated visual opens with the message **This visual was generated by AI. Review every layer before taking it on air.** Check the texts, link them to [live data](https://docs.visualradioassist.live/visual-designer/insert-data.md) where needed, and save. For a graphic from your branding, **Try another** generates a new version.

## AI credits for generated graphics

Each generated graphic costs 1 AI credit, charged only when the generation completes. A failed generation costs nothing. Contact support to check how many credits a station has left.

* A station generates at most 10 graphics per day, failed ones included.
* A station generates one graphic at a time.
* When the credits for the month are used, generating stops until they renew. Contact support for more.

Importing your branding from your website costs no AI credits.

## Related

* [Set up your station branding](https://docs.visualradioassist.live/visual-designer/branding.md)
* [Visual templates](https://docs.visualradioassist.live/visual-designer/visual-templates.md)
* [Import from Photoshop](https://docs.visualradioassist.live/visual-designer/import-from-photoshop.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.