# Adobe Firefly Gets a Major AI Boost with New Image and Video Models, Redesigned Web App

## Adobe Firefly Gets a Major AI Boost with New Image and Video Models, Redesigned Web App

Adobe has unveiled a significant upgrade to its Firefly AI platform, launching new image and video generation models, along with a revamped web app that integrates AI capabilities from competitors. The move solidifies Adobe’s commitment to generative AI and positions Firefly as a central hub for creative AI workflows.

The core of the update is the **Firefly Image Model 4**, promising enhanced quality, speed, and control over image generation. Users can now fine-tune outputs with greater precision, dictating camera angles, zoom levels, and the overall style. Notably, the model can generate images at resolutions up to 2K. For even more demanding tasks, Adobe is introducing **Image Model 4 Ultra**, a souped-up version capable of rendering complex scenes with intricate details.

According to Alexandru Costin, VP of Generative AI at Adobe, the new models were trained using significantly more computational power, resulting in more detailed and realistic images. He also highlighted improvements in text generation within images and the ability for users to leverage their own images to guide the model’s stylistic output.

The **Firefly video model**, previously in limited beta, is now available to all users. This tool allows the creation of video clips from text prompts or images, offering control over camera angles, start and end frames, atmospheric elements, and motion design customization. The video model can generate clips from text at resolutions up to 1080p.

Beyond images and video, Adobe is also introducing the **Firefly Vector Model**, designed for creating editable vector-based artwork. This model can be used to generate logos, product packaging, icons, scenes, and patterns, offering a new avenue for graphic design and illustration.

Perhaps the most intriguing aspect of the update is the redesigned Firefly web app. It not only provides access to Adobe’s own AI models but also integrates image and video generation capabilities from **OpenAI (GPT image generation), Google (Imagen 3 and Veo 2), and Flux (Flux 1.1 Pro)**. This unified platform allows users to seamlessly switch between different AI models, all while ensuring that generated images have content credentials attached for transparency. Adobe has hinted at the possibility of adding even more AI models in the future.

Adding to its suite of creative tools, Adobe is also publicly testing **Firefly Boards**, a canvas for ideation and moodboarding. This collaborative space allows users to generate or import images, remix them, and work with others, similar to platforms like Visual Electric, Cove, and Kosmik. Firefly Boards is accessible via the Firefly web app.

The company hasn’t provided a specific timeline, but these new models are expected to be integrated into Adobe’s existing product portfolio soon.

Furthermore, Adobe is making its **Text-to-Image API and Avatar API generally available**, and a new **Text-to-Video API is now available in beta**. These APIs can be accessed through the company’s Firefly Services collection, offering developers and businesses the opportunity to integrate Firefly’s AI capabilities into their own applications and workflows.

Finally, Adobe is testing a web app called **Adobe Content Authenticity**, allowing users to attach credentials to their work to verify ownership and attribution through metadata. Users can also specify whether their images can be used by AI companies for training purposes.

With this comprehensive update, Adobe is positioning Firefly as a powerful and versatile platform for generative AI, catering to a wide range of creative professionals and industries. The integration of third-party AI models and the focus on content authenticity further solidify Adobe’s commitment to responsible and innovative AI development.