Adobe presents its new AI assistant Firefly, capable of transforming words into creative projects

Adobe has presented his new artificial intelligence (AI) assistant Firefly that allows you to create content in Adobe Creative Cloud applications simply through directions with words in a conversational interface in the Firefly app as it runs complete multi-step workflows.

The company has expressed its intention to bring the full capabilities of your content creation and editing tools to creators of all levels automating and simplifying the startup process and allowing you to move quickly “without having to navigate multiple applications or perform manual steps.”

In this sense, it has presented its new AI assistant Firefly, that allows you to create multimedia content with just describe the desired result through words in a conversational interface.

As explained in a statement, this assistant is capable of generating the content requested by orchestrate and execute complex multi-step workflows in Adobe Creative Cloud applicationsincluding Firefly, Photoshop, Premiere, Lightroom, Express, and Illustrator, among others.

To do this, it mixes conversational and generative assistance AI, and is able to understand the content being created, including images, videos, designs and brand elements, to carry out more relevant actions adapted to the context.

All of this is carried out in a unified conversational interface within the Firefly appwhere the assistant displays workflows, results, and maintains context, progress, and decisions between sessions. Likewise, it must be taken into account that this context can be transferred easily at individual Adobe applications to continue developing the project.

Additionally, the wizard also helps users with their creations asking contextual questionsfacilitating decision making and offering suggestions. However, you can intervene in the process at any time and guide, refine or adjust the results.

The Firefly AI assistant also features a library of predefined creative skills Designed specifically for creative workflows, they allow the assistant to execute complex, multi-step tasks from a single prompt. Thus, users can Customize these skills and create your own so that the assistant automatically performs tasks according to your taste, for example, when retouching portraits with presets.

Added to all this is that the assistant can learn from user preferences over timeincluding the preferred aesthetic options, offering results that are more tailored to your needs automatically.

Finally, for shared workflows, creators can ask the assistant to organize and share their work on Frame.iowhere interested parties can review it and provide feedback. After that, the assistant It will collect the comments, interpret them and automatically apply the changes.

This is “a fundamental change in the way creative work is done”, as Adobe has stressed, since it allows creators simply guide the assistant “saving time and effort” and eliminating the barrier that the technical use of these tools represents on some occasions.

However, the company has clarified that the control of the created material remains with the user in questionsince they provide “the judgment and creative direction”, while the assistant is in charge of the organization and execution.

Furthermore, it has highlighted its use also for creative professionals, allowing drive more complex multi-step workflows, “combining speed with the control and precision necessary for high-quality creative work.”

However, Adobe has shared that will also bring this way of creating content to major third-party AI modelslike Claude de Anthropicallowing users to access Adobe tools and its assistant directly in these applications.

NEW VIDEO EDITING CAPABILITIES AND MODELS

In addition to all this, Adobe has also expanded video and image editing with AI. Specifically, it has introduced new features in Firefly Video Editor, such as studio-quality sound, advanced color adjustments and integration with the Adobe Stock image bank.

It has also added new precision image editing capabilities, such as tools Precision Flow y AI Markup.

Another novelty is that it has expanded its catalog with more than 30 creative AI models, giving creators greater control to choose the desired models. Thus, they now have options such as Kling 3.0 and Kling 3.0 Omni, which join Nano Banana 2 and Veo 3.1 from Google, Gen-4.5 from Runway and Multilingual v2 from ElevenLabs, among others.