A prompt is a short descriptor of text that the Scenario Bot interprets using AI Models to produce an image.
The Model analyzes the words and phrases used in a prompt and compares the results to its own training and then uses that information to generate an image.
The art of Prompt-crafting can open many doorways to unique ideas and exciting iterations!
To prompt the Scenario Bot, simply type /create
Select /create from the popup by clicking it or simply press the space bar:
Type in your prompts separated by a comma ,
Press Enter and in a few moments you will see your results:
Negative Prompts
It is often useful to prompt something out of an image.
Negative prompts work the same as regular prompts but instead of telling the Model what you want to create, you are telling the model what not to create.
After typing /create in the message bar and selecting it, you’ll notice that next to the prompting field there are words in gray that read ‘+1 more’:
Click ‘+1 more’ and select ‘negative’ to open the Negative prompting field
you may now add negative prompts to your image generation
Prompting Basics
A basic prompt can be a simple word or phrase.
A good way to think about basic prompts is the way you would use hashtags on twitter or Instagram. They are descriptors that can refer to subjects, objects, nouns, adjectives, styles, hardware, etc. that you want to appear in the image output.
Prompts closer to the beginning of the line will carry more subject weight
This means that prompts describing an action or a motion and a main subject, i.e. ones that ‘set the scene’ will have more weight when placed at the beginning. Prompts closer to the end of the line will carry more style weight; this means descriptors such as ‘anime’, ‘gothic’, ‘cyberpunk’, etc. will have more weight when placed at the end.
Test Prompting
Prompt engineering is an important skillset in generative art creation.
A helpful technique for testing a Model is to create a list of prompts that cover a range of different subjects, styles, and compositions. This will allow you to get a sense of how your model handles different types of input, and identify any areas where it may struggle or perform poorly.
Prompt Expressions
For more information on prompting, learn about our Prompt Expressions by clicking here