# Images, music, voice and slides

> The media the agent makes for your project besides code - images, video clips, music, sound effects, narration and presentations - with costs and limits.
> Source: https://playcode.io/docs/build/images-music-voice-and-slides - last reviewed 2026-10-09.
> Ask Playcode's agent: "Generate a logo for Bloom Bakery and put it in the site's header." It generates the image, saves it in the project and adds it to the page; generation needs a Pro or Max plan and uses credits.

Besides code, the agent makes the media a project needs: images, short video clips, music, sound effects, spoken narration and presentations. It saves each file into your project, so you can ask for an asset and see it on your page in the same conversation.

## What the agent can make

| Media | What you can ask for | Plan |
|---|---|---|
| Images | Logos, icons, photos and illustrations; changes to a picture you attach; a transparent background for a logo or an icon | Pro or Max |
| Picture processing | Resize, rotate, compress or convert pictures you already have, to WebP, AVIF, PNG, JPEG, GIF or a favicon | Every plan |
| Video clips | A clip from a description or from a picture: 4 to 30 seconds long, landscape (16:9) or portrait (9:16), up to 4K, silent unless you ask for sound | Pro or Max |
| Video processing | Trim, resize, compress, speed up or slow down a video, remove its sound, or convert it to MP4 or WebM | Every plan |
| Composed videos | A video built from your text, screenshots and music, rendered as an MP4 in any frame size, square included | A Cloud app |
| Music | A background track from 3 seconds to 10 minutes, cut to the exact length you need, with or without vocals | Pro or Max |
| Sound effects | A click, a whoosh or an ambience from 0.5 to 30 seconds, looping if you want | Pro or Max |
| Narration | Spoken voice from your script, up to 2,500 characters per part. A longer script is read in parts | Pro or Max |
| Presentations | A slide deck you open by a link and present in the browser, a PowerPoint file (.pptx) you download and edit, or both | Every plan. A PowerPoint file needs a Cloud app |

On the Free plan, the agent says that generation needs a paid plan and continues with the rest of your request.

## What it costs

Each image, clip, music track and narration uses credits from the workspace's balance, on top of the agent's own work. Video costs the most, because it is charged per second. **Settings**, **Usage** shows what each one cost.

A new generation gives a new result, not a better take of the last one. A clear, detailed description the first time saves credits. A composed video has no per-second charge: it uses the time of the Cloud app's computer.

## What the agent can check

- **Images:** it looks at each image it makes before it uses it.
- **Video:** it cannot watch video. It asks for a text description of what is on screen, and checks the length and the size.
- **Audio:** it cannot hear. It checks the length and the loudness, so listen yourself before you publish. It can mispronounce names and abbreviations.

## On your site

The agent saves each file into the project and adds it to the page you asked for. Browsers do not play sound when a page opens, so the agent starts audio from a click, or adds video without sound and a button to turn it on.

A generated clip and a music track stay two files: the agent can build a page that plays them together. For one video file with the music in it, ask for a composed video in a Cloud app.

## Presentations

A deck you open by a link works in a static site and in a Cloud app. A PowerPoint file needs a Cloud app, because a server writes the file: in a static site, the agent first offers to convert it and says why. The PowerPoint file has real text boxes you can edit, and it can look slightly different in each program that opens it. There is no PDF or Keynote export.

You provide the facts and figures in a deck, and you check them before you present.

## Limits

- Images are picture files, not editable vector files. A generated logo is not checked against trademarks.
- Video clips are 4 to 30 seconds. Keep a composed video short: the agent aims for 10 to 15 seconds, because the render runs on the project's computer.
- Generated music works best as a background track. Vocals are its weakest part.
- Narration uses a default narrator voice. It cannot copy a real person's voice.