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

# How do I make images and videos with AI Studio?

> Make images and short videos with AI in AI Studio: the models, shapes and lengths you can pick, what each costs in credits, and where your results are saved.

**Plans:** Pro and Business, and the free trial. One image model, Nano Banana Pro, is Business only.

AI Studio makes images and short videos for your posts. It shows the price in credits before you generate, and saves the results to your **Media Library**.

## Two ways to start

Open **AI Studio** in the sidebar. At the top, choose how you want to work:

* **Describe it.** Type what you want to post, the way you'd text it. Click **Set up my post**. AI Studio works out the post type, the layout and the wording. This step is free. Check what it filled in, then generate.
* **Set it up myself.** Pick **Image** or **Video**, then a post type, and fill in the details.

### Post types

Under **What do you want to post?** you'll find ready-made post types, like **Sale or discount**, **Product spotlight**, **Event announcement**, **Testimonial** or **Quote card**. Each asks a few simple questions and picks a good model for you.

Post types also write a caption to go with the image or video. That adds 1 credit to the price. To write the caption in your brand's voice, pick a kit under **Brand**. The kit shapes the caption only, not the image or video.

To pick the model and write the whole prompt yourself, choose **Start from scratch**.

## Make an image

<Steps>
  <Step title="Pick Image and a post type">
    Or pick **Start from scratch**.
  </Step>

  <Step title="Choose a model">
    Open the **Model** list. Under it you'll see the price, like "14 credits per image".
  </Step>

  <Step title="Pick an aspect ratio">
    Every image model offers 1:1, 4:3, 16:9, 3:4 and 9:16.
  </Step>

  <Step title="Write your prompt">
    Describe the image in the **Prompt** box, up to 4,000 characters.
  </Step>

  <Step title="Add a reference image if you like">
    Some models can follow an image you give them, under **Reference files**. Upload one, or click **Pick from media library**.
  </Step>

  <Step title="Check the price and click Generate">
    **Est. cost** shows the credits it will use.
  </Step>
</Steps>

### Image models

| Model | Credits per image | Reference images |
| - | - | - |
| FLUX.2 Draft | 4 | No |
| Seedream v4 | 8 | No |
| Nano Banana | 11 | Optional, up to 4 |
| Seedream v4.5 | 11 | Optional, up to 4 |
| Flux Kontext Pro | 11 | Needs 1. It edits the image you give it. |
| GPT Image 2 | 14 | No |
| Ideogram V3 | 16 | No |
| Nano Banana 2 | 21 | Optional, up to 4 |
| Nano Banana Pro (Business) | 40 | Optional, up to 4 |

Each generation makes one image. Reference files can be up to 50 MB each.

## Make a video

Pick **Video**, then choose a model, an **Aspect ratio** and a **Duration**. Write your prompt and click **Generate video**. Videos usually take 30 to 60 seconds.

| Model | Aspect ratios | Length and credits |
| - | - | - |
| Kling 2.5 Turbo Pro | 1:1, 16:9, 9:16 | 5s: 92, 10s: 184 |
| Wan 2.5 | 1:1, 16:9, 9:16 | 5s: 132, 10s: 263 |
| Veo 3.1 Fast, with sound | 16:9, 9:16 | 4s: 158, 6s: 237, 8s: 315 |
| Veo 3, with sound | 16:9, 9:16 | 4s: 420, 6s: 630, 8s: 840 |

**Turn an image into a video.** Kling 2.5 Turbo Pro, Wan 2.5 and Veo 3.1 Fast each have an "animate image" version. Give it one image and describe the motion. The video takes the shape of your image and costs the same as above.

<Note>
  The prices here are the ones AI Studio showed when this page was written. The price on the button is always the one you pay.
</Note>

## Not enough credits?

You'll see "You need 12 more credits to generate this." with your own number, and a **Buy credits** button. Pick a cheaper model, or [buy more credits](/help/ai/buy-credits).

## While it's working

AI Studio makes up to 3 things at a time for each workspace. Anything more waits as **Queued…**, and you can **Cancel** it then at no cost.

When it's done, you'll see "Your image is ready." or "Your video is ready." You also get a notice in the app.

## If it fails

You'll see "Generation failed - credits refunded." The failed card says how many credits came back. A generation that takes more than 30 minutes fails and is refunded too.

<Warning>
  If the model's safety filter blocks your prompt, you may not get the credits back, even if the message says so. Check the refund line on the failed card. If no credits came back, email [hello@publishbuddy.com](mailto:hello@publishbuddy.com).
</Warning>

## Where your results go

Everything you make shows in the history on the right. Use the filters to see **Images**, **Videos** or **Failed** ones. Everyone in the workspace sees the same history.

Each result is also saved in the **AI Studio** folder of your **Media Library**.

Click a result to open it. From there you can:

* **Use in new post:** opens the composer with the image or video and its caption.
* **Regenerate:** fills in the same settings so you can change them and try again. It doesn't charge until you generate.
* **Download:** opens the file so you can save it.
* **Delete:** removes it from AI Studio and your Media Library.

## Related

* [How do AI credits work?](/help/ai/ai-credits)
* [Make a carousel with AI](/help/ai/ai-studio-carousels)
* [Brand Kits](/help/ai/brand-kits)
* [Use library media in a post](/help/media/use-library-media)


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