Dynamic Prompts
Templating language and Python library for generating prompts for text-to-image generators.
Create flexible templating workflows and combinatorial prompt variations in Python.
Dynamic Prompts is a Python library providing an intuitive templating language for generating text-to-image prompts. It is ideal for developers and AI creators wanting to automate and scale prompt generation workflows.
What it is
The library lets you create unique prompts from a single template using variants, wildcard files, and variable assignment. It supports exhaustive combinatorial generation, Jinja2 templates, and helper features like Magic Prompt and Lexica.art search.
Who it's for
It suits Python developers and digital artists building automated pipelines for Stable Diffusion or similar generators. It does not suit users looking for a standalone no-code application without writing code or managing wildcard files.
Key features
Wildcard File Support
Use text, JSON, or YAML files as placeholders in your templates.
Combinatorial Generation
Exhaustively generate every possible prompt combination from a given string template.
Jinja2 Templates
Leverage Jinja-powered templating for advanced and complex prompt creation.
Attention Grabber
Emphasize random phrases in your prompt for systems supporting attention syntax.
Up and running in 3 steps
Our honest verdict
What's great
- Supports multiple wildcard file formats including text, JSON, and YAML.
- Includes combinatorial generation to enumerate every possible prompt variation.
- Provides Magic Prompt integration to automatically expand and spice up prompts.
- Supports Jinja2-powered templates for advanced prompt engineering logic.
Where it falls short
- Requires writing Python code or integrating with specific pipelines like Automatic1111.
- Combinatorial generation can quickly produce thousands of unintended prompts if not bounded.
- Additional features require installing extra package dependencies.
Pricing & license
Free under the MIT license with no paid hosted option; the only cost is running it yourself within your Python environment.
Open remote AI-training roles
873 open remote AI-training roles, updated November 15, 2026; most pay $20–$170 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
Use it, or have us set it up
FAQ
How do I use wildcards?
You can initialize a WildcardManager pointing to your directory of text, JSON, or YAML files and reference them using double underscores like __colours__.
Can I generate every possible combination?
Yes, using the CombinatorialPromptGenerator will produce every possible prompt from your template up to a maximum limit.
Does it support random seeds?
Yes, you can pass a single random seed, a list of seeds matching the prompt count, or an int value for reproducible outputs.
