Adobe Firefly, in a nutshell, serves as Adobe’s take on the rising artificial intelligence trend. It’s a host of feature-packed image-generation software that tackles digital images, illustrations, vectors, videos, social media publication materials, and 3D modeling.

The company promises that the AI will only be trained on Adobe Stock, openly licensed content, and public domain material with expired copyright. Plus, creators of Adobe Firefly have the freedom to decide whether or not they’ll allow the software to use their graphics and designs for training. Currently, the software is released as a demo with only two features available.

Generate images or text textures with a prompt

As it is in its earliest stage of development, Adobe Firefly only has two features available for users to try — which are converting text into images and adding styles to typed-out text. The former is reminiscent of Midjourney, and you only need to type in a few words to generate an image. This resulting photo can be further modified with filters, adjusted lighting, and content type.

The latter feature focuses on the appearance of a typeset. Imagine you want a host of letters with a texture that looks like melted chocolate, type in the prompt, and voilà — that’s what you’re going to get. You can even change the font of the resulting text. Meanwhile, the upcoming features will enable you to edit images and videos with greater detail, generate custom brushes and vectors, and more.

Currently, Adobe is highly encouraging interested users to join the beta testing stage. In time, it plans to give users virtually unlimited access to explore the boundaries of their creativity. Users can also let the software use their work in its “training” or use the “Do Not Train” tag on their content. However, like with any other AI-powered image-generation software, this one is also a hit-or-miss in its results.

Create and do more with its recent features drop

Adobe Firefly has rolled out some cool new features in its latest updates. Users can now create, share, and collaborate on ideas in real-time, make videos with transparent backgrounds, and whip up vector images using partner models like GPT Image. The platform also lets you fine-tune video generation with advanced controls, resize multiple images while focusing on important parts, and generate content that’s safe for commercial use right from your phone.

The new Firefly Boards (beta) really step up the creative game by making it super easy to organize and connect your videos and images. You can also create sound effects just by typing prompts or even using your own voice, which makes your projects sound way cooler. Plus, with options like virtual avatars for video making and auto-reframing for different aspect ratios, Firefly gives you awesome tools that work great for all kinds of content creators.

A careful approach to AI image generation

Overall, Adobe Firefly is a careful step into the controversial world of AI-powered image generation. Adobe isn’t reckless in its execution, seeing as the software is currently trained on the company’s existing royalty-free media library, openly licensed content, and material with expired copyright. Although the software is still in its infancy and two features have been revealed. Plus, it also commits common AI errors like generating bad results.