The Evolution of Project Dream

Introduction

Since its initial release, we’ve worked hard to deliver new features, models, and, most importantly, learn from you. As an AI visualization tool, Project Dream continues to evolve through your real-world feedback.

Everyone thinks about AI from a different point of view, so their approaches and expectations vary widely. Creating images with AI is still a highly iterative process and takes time and effort to learn how to generate exactly what you’ve imagined, whether for image-generation archviz projects or broader creative needs. Learning the different settings and using the tools available will help improve your chances of achieving that.

It’s all about getting closer and closer...

TL;DR

Project Dream is evolving through real-world user feedback, recognizing that AI image creation is iterative and user needs vary. Recent updates (0.5.1–0.5.5) include:

• Ratings and public images to guide defaults and help newcomers

• A Faithful input method with a Creativity slider for controlled fidelity

• New Artistic and Flexible models

• Compare Image mode for side-by-side evaluation

• Major speed boosts with AI render upscaling to 4K

• Reference mood image input for guiding color and tone

• A reworked result viewer with zoom and clipboard copy

• Prompt-editable Variations for larger creative shifts

These improvements streamline experimentation, enhance control over fidelity and style, and boost quality and speed. The team invites community participation via Discord and encourages trying the free trial.

Release Notes

Version 0.5.1

• Added the ability to rate your generations and make them public.

• Ratings help us understand which styles, input methods, and settings are most effective, allowing us to dial in better default settings.

• Public images help newcomers see what prompts and settings lead to great results.

Version 0.5.2

• Introduced a new input image method called Faithful.

• Renamed Image input strength to Creativity for clarity.

• Faithful uses your input image as-is and varies embellishment based on the Creativity slider:

  • Creativity 0: minimal changes; fills in details.

  • Creativity 100: more freedom to add details and “funky” elements.

Version 0.5.3

• Added two new models: Artistic and Flexible.

• Enabled copying your job details entirely back into the input field for faster iteration.

• Introduced Compare Image mode to intuitively compare your input image with generated results.

• Optimized performance for a massive speed boost and enabled AI render upscaling up to 4K.

Version 0.5.4

• Added a reference mood image input to influence mood and color palette.

• Released a completely reworked result viewer with easy zoom for upscales.

• Added the ability to copy generated images to the clipboard for streamlined workflows.

Version 0.5.5

• Enabled changing prompts when creating Variations jobs.

• Allows using vaguer prompts to make bigger creative changes to images you already like.

More to come soon!

Community and Getting Started

Come and join us in our Discord channel, and let us know how you are using Project Dream, or what you really miss from it!

Check out Project Dream for yourself, and see what kind of inspirations you end up crafting.