A simple and compact batch prompt workflow for Flux2 Klein, based on the workflow uploaded on reddit by user RIP26770.

The function of this workflow in how it is setup is to generate a batch of different viewing angles from an input image. Each line is its own prompt that will generate an individual image. It's as simple as loading an input image and clicking run. When all prompts are completed, an interactable gallery of the outputs will appear.

There will obviously be limitations to this approach, as it's limited by the model's understanding of things, but it can perform pretty well. Minor adjustments to prompts can be useful, such as specifying a white or black background or any other useful change. It can also be used to, for example, generate a series of different style alterations to an input image in one go. So pretty much any procedural type of prompting strategy is a good use case for this workflow.

_________

Required extension: https://github.com/ai-joe-git/ComfyUI-Simple-Prompt-Batcher

_________