---
title: Stable Diffusion
description: "Stable Diffusion generates images through the **Stability AI** API. You must provide your own API key. Generation usually takes **4 to 12 seconds**. **Pro.** This bank is locked on the Free plan (Pro badge on Source). Source badge: **AI**. **Gutenberg:**…"
url: "https://magic-post-thumbnail.com/docs/stable-diffusion/"
updated: "2026-08-16"
category: Image Databases
---

# Stable Diffusion

Stable Diffusion generates images through the **Stability AI** API. You must provide your own API key. Generation usually takes **4 to 12 seconds**.

**Pro.** This bank is locked on the Free plan (Pro badge on Source).

Source badge: **AI**.

**Gutenberg:** not available as a block search tab. Photo-bank tabs in Gutenberg are Openverse, Flickr, Pixabay, and (Pro) YouTube, Unsplash and Pexels. Google Image (API) and AI generators are Source / automatic generation only.

## Open the bank in Source

In version 7, banks are configured from **Magic Post Thumbnail → Source**. Click the bank in the left column (or tick it to add it to the chain), then click **Configure**. The **Banks Settings** panel opens on the right. Save when you are done.

The left column is the **Automatic generation chain** used by the editor button, bulk generation, cron and imports. Drag to set priority: the first bank is tried first. Unchecked banks sit below the chain.

![Stable Diffusion bank settings](https://magic-post-thumbnail.com/wp-content/uploads/2026/08/panel-stability.png)

## Get an API key

1. Create a key at [https://platform.stability.ai/account/keys](https://platform.stability.ai/account/keys).
2. Paste it into the Source panel, click **API testing**, then Save.

Style presets from Automatic still apply to this AI bank.

Source Preview a search can consume credits on AI banks. Use a cheap model while testing.

## Model

Default: **Stable Image Core**.

- **Stable Image Core** — fast, recommended. Built for websites and bulk. Fastest / cheapest of the current Image models for volume.
- **Stable Image Ultra** — best quality. Photorealistic; uses more credits.
- **SD 3.5 Large Turbo**
- **SD 3.5 Flash** — cheapest SD 3.5 option.
- **SD 3.5 Large**
- **SD 3.5 Medium**

## Other options

- **Aspect Ratio** — 16:9 (default), 3:2, 1:1, 4:5, 5:4, 2:3, 9:16, 21:9, 9:21.
- **Style** — None, Photographic (default — best for featured images), Cinematic, Digital art, Enhance, Anime.
- **Output Format** — JPEG (default), PNG, WebP.