Sign In

Random photorealistic prompt

Updated: Feb 8, 2026

style

Download

1 variant available

Archive Other

10.4 KB

Verified:

Type

Workflows

Stats

57

Reviews

Published

Feb 8, 2026

Base Model

SDXL 1.0

Hash

AutoV2
7F25A4BC25

This was inspired by PromptGeek's YouTube video and eBook so all credits goes to them. Please go and show them some love: Creating Photorealistic Images With AI: Using Stable Diffusion

This workflow builds a random prompt with keywords from the eBook mentioned above following this pattern (also from the eBook): [STYLE OF PHOTO] photo of a [SUBJECT], [IMPORTANT FEATURE], [MORE DETAILS], [POSE OR ACTION], [FRAMING], [SETTING/BACKGROUND], [LIGHTING], [CAMERA ANGLE], [CAMERA PROPERTIES],in style of [PHOTOGRAPHER]

In the "Inputs here" group, you set a seed, select a checkpoint and image size. To the right, you set your subject, and optionally a negative prompt. The generated positive prompt can be seen in the console.

In the "Random parameters" group, you can add and remove keywords. You may also want to edit the output path.

I use two custom node packs. The first one, was-node-suite-comfyui, is a must. It's apparently retired but works fine: https://github.com/WASasquatch/was-node-suite-comfyui

The other one, cg-use-everywhere, just keeps things neat. You can connect seed, model, clip, vae, and latent manually if you don't want to install it: https://github.com/chrisgoringe/cg-use-everywhere

Have fun.