# Run SDXL Turbo

Draws an image in about a second. Built for trying ideas quickly. Ready in about 90 seconds on an RTX 3090, from $0.22/hr. No install, no code.

## What it does

SDXL Turbo is a stripped-down version of Stable Diffusion XL that produces an image in roughly one second instead of eight.

The trade is quality. Turbo images are looser and less detailed, so this is a sketching tool rather than a finishing tool.

The usual way to work is to explore with Turbo until you find a prompt you like, then run that same prompt through Stable Diffusion XL or FLUX for the final version.

## Your setup

We run SDXL Turbo on an RTX 3090 with 24GB of memory, at $0.22 per hour, billed by the minute. It takes about 1 second per image.

Turbo uses a single denoising step where standard Stable Diffusion XL uses around thirty. That makes it memory-light and almost entirely limited by how fast the card can move data, not by raw compute. The RTX 3090 has the memory bandwidth to keep up at half the hourly price of a 4090, so the cheaper card is genuinely the better match here.

## What it costs

1,000 images takes about 20 minutes of GPU time, which is about $0.07.

## Two ways to run it

- **Open the app.** A live workspace that redraws as you type. Good for finding a direction fast.
- **Run a batch job.** Bulk-generate variations from a prompt list when you need many options at once.

## Model facts

| Field | Value |
|---|---|
| Parameters | 3.5B |
| License | Stability AI Non-Commercial Research Community License |
| Memory required | 8GB minimum |
| Base model | Distilled from Stable Diffusion XL |
| Publisher | Stability AI |
| Recommended hardware | RTX 3090, 24GB |
| Category | [Image generation](https://gpuvault.io/working-on/image-generation/) |

## Questions people ask

### Do I need to install anything to use SDXL Turbo?

No. We start a machine with SDXL Turbo already loaded and hand you a link. Everything runs in your browser, there is nothing to download, and nothing is left on your computer afterwards. It works the same on a Mac, a Windows laptop, or a Chromebook. A workspace is usually ready in about 90 seconds.

### How much does it cost to run 1,000 images?

About $0.07. SDXL Turbo takes about 1 second per image on the RTX 3090 we recommend, so 1,000 images is roughly 20 minutes of GPU time at $0.22 per hour. Billing is by the minute, and a batch job shuts the GPU off the moment the last item finishes. The calculator above works this out for your own numbers.

### Can I run SDXL Turbo on a cheaper card?

Sometimes, and the calculator will not always make it look worth it. Turbo uses a single denoising step where standard Stable Diffusion XL uses around thirty. That makes it memory-light and almost entirely limited by how fast the card can move data, not by raw compute. The RTX 3090 has the memory bandwidth to keep up at half the hourly price of a 4090, so the cheaper card is genuinely the better match here. If you want to try a different card anyway, the advanced catalog lets you pick one and shows the estimated time before you commit.

### What are the license restrictions on SDXL Turbo?

SDXL Turbo is released under the Stability AI Non-Commercial Research Community License. That is not a permissive license, so read it before you use the output in paid work. We show the license on every model page precisely because this catches people out, and there is usually a permissively licensed alternative in the same category.

### What happens to my files after I stop?

Anything you save into your workspace folder stays in your account and is there when you come back. Everything else is destroyed: the machine is terminated and its disk is wiped. Batch job inputs and outputs are deleted from our storage 24 hours after you download them. We do not read your files or use them to train anything.

### What if it fails partway through?

You are not charged for work that did not complete, and failed sessions are refunded automatically. Batch jobs write each output as it finishes rather than at the end, so a failure at item 380 of 400 leaves you 379 usable files and a message saying what broke. Restarting picks up where it stopped rather than redoing the work.

---

*Source: https://gpuvault.io/model/sdxl-turbo/ — GPUVault, GPU rental by the minute.*